[TowerTalk] Ferrites 31 vs. 77 material

jimlux jimlux at earthlink.net
Sat Nov 9 13:53:32 EST 2019


On 11/9/19 10:32 AM, Jim Brown wrote:
> On 11/9/2019 7:45 AM, jimlux wrote:
>> Big cores (high mu) with lots of turns (big N^2) is the way to go..
>> (Except that at high frequencies, the capacitance between the turns 
>> short circuits the impedance of the windings - so there *is* a happy 
>> medium)
> 
> Jim,
> 
> It's FAR more complex than that. Common mode chokes work by using the 
> RESISTANCE coupled from the core, in the region where the choke is 
> resonant. The ideal core material has a very low Q, providing a broad 
> resonance and thus a wide effective bandwidth. #31 is a well engineered 
> material for HF -- it is MnZn material that combines the circuit 
> resonance (the turns) with it's dimensional resonance, which serves to 
> broaden the resonance in the HF spectrum.
> 
> BTW -- because R is coupled from the core, it also increases as N 
> squared, and C increases with N.


Yes, it is a gross oversimplification, but I think it *is* a short 
explanation why many turns on a bigger core is better than a string of 
beads, particularly at lower frequencies.






> 
> These concepts are developed in k9yc.com/RFI-Ham.pdf  which is based on 
> work I published in an AES Paper in 2005. I learned of dimensional 
> resonance from a classic work by Snelling, which is referenced in the 
> paper and the tutorial. A colleague who was on faculty at the Univ of 
> Chicago at the time found it in the sub-basement of their engineering 
> library and loaned it to me. :)  That AES paper is also on my website -- 
> scroll way down.


yes, that Snelling book is hard to come by. I got a copy at work, where 
they borrowed it from Armstrong Flight Research Center.


> 
> There are a number of MnZn materials, but #31 is the only one that is 
> useful from 160-6M. #75 is useful for transmitting chokes on 630M, and 
> smaller #75 cores are useful for RX chokes.
> 
> #77 and #78 materials are NOT useful for chokes because their 
> dimensional resonance in the range of 1 MHz is VERY high Q -- in other 
> words, it does NOT combine in a useful manner with the circuit resonance.
> 


More information about the TowerTalk mailing list