Topband: Binocular Cores

[Tom W8JI]

> That's good info, if you understand how to interpret it. It's not clear
> what u' amd u" are. And the info on their site doesn't tell you just how
> good (low losses and reactance across a wide bandwidth, among other 
> things)
> a matching transformer is made from a 73 material binocular core. They 
> only
> offer it as RFI suppression.

Mike,

You might want to go to this page 
http://www.fair-rite.com/newfair/index.htm
and hover the link button [technical]>[Use of Ferrites in Broadband 
Transformers]

For receiving or low power transmitting, where voltage across a winding is 
minimal, you can use a core that presents a resistive impedance. That shunt 
impedance parallels the windings, so it has to be high compared to the 
operating impedance or the transformer will be lossy. Receiving or low power 
applications are very forgiving.

For high power, we get into trouble quickly. Since voltage across a winding 
can be fairly high, the parallel resistance has to be very high. Even then 
we can get into trouble. If a core dissipates just a few watts of heat it 
can go into thermal runaway, where it quickly reaches Curie temperature and 
no longer acts like a magnetic core.

Think about choke baluns where they have 10,000 ohms impedance, dominantly 
resistive. If you applied a 500 ohm load and 1000 watts you would have 2000 
V/2 =1000 volts in an ideal balance condition, which is 100 watts in the 
cores. This is why core selection generally has to shift to a dominantly 
reactive impedance, rather than resistive, at high power.

With a 100 watt lower voltage PA we might be able to use a  52 material. If 
it is higher voltage across the winding, the shunting resistance might kill 
it with heat. With high voltages across windings we might have to drop from 
a  52 material to 61, just to stop long term heating while maintaining the 
high end response.

What is simple and unimportant for low winding voltages really becomes a 
much more complex series of compromises at higher voltages across windings. 
That parallel resistance caused by the core is more problematic for heat 
than it is for loss.

73 Tom