FSL Experiments


Robert Conboy
 

FSL builders, I have some questions…

 

(but first, sorry for the big font)

How is the selectivity of your antennas? Is it significantly narrower than 10 kHz? i.e. when tuned spot-on, does audio become muffled because the treble is greatly reduced? This is desirable because it lets you tune to one sideband or the other.  

 

I find this to be the case at lower frequencies but less so at higher frequencies on some of the antennas I have made. 

 

I have found it helps to use a ganged capacitor and connect the coil between the two stators. When doing so, inductance needs to be doubled, which is about 1.5 times as many turns. Lead dress becomes more important and lead lengths must be as short as possible (you might need to offset the coil toward one end of the ferrite). The Litz also becomes more important. Aside from raising the Q, it makes the capacitor much less sensitive to hand capacitance.  

 

How does this raise Q?

If you connect your coils between the two stators of a ganged capacitor, you take the contact resistance to the capacitor’s frame out of the circuit, improving the Q. But perhaps more significantly, doubling the reactance (twice the inductance/half the capacitance) while only increasing the coil resistance by 50 % increases the Q.  Furthermore, ferrite losses become less significant  in proportion to the increased coil/capacitor reactance.

 

 

My latest FSL build uses a 5-500 pF vacuum variable capacitor and 250uH of  3000/46 Litz on 49 ferrite rods mix 61, 200mm x 10mm, roughly 7 inch diameter FSL.

 

I have determined that using such exotic L and C components improves Q very little over what you get with a conventional capacitor and 500-ish to 1000-ish strands of Litz. There is an inflection point beyond which Q-degrading losses in the ferrite become dominant. 

 

To achieve values of Q that I want, I use regenerative feedback to overcome the ferrite losses. It works very well. I use a 500 Ohm ten-turn pot to adjust the feedback, in series with a parallel combination of 150 Ohms and 1360pF. A mini-circuits ZFL-1000+ supplies the gain. The amp is powered by 12 AA nimh batteries (14.4 Volts) and the batteries run about 30 hours per charge. The pickup coil is a shielded loop of RG174 located 2 inches off the end of the ferrite sleeve. The electronics are in a 6 inch diameter plastic jar glued to the end of the FSL former (6 inch foam roller) controls are mounted on the jar’s lid.

The feedback coil is a single turn of 1/2 inch copper tape stuck to the ferrite.  The feedback coil and batteries can be disconnected via a dpdt switch.

 

For the rest of us who don’t know the technicals:

 

In a vacuum variable capacitor, both sets of “plates” are each a series of concentric cylinders. One set is fixed rigidly and remains stationary. The other set is linearly movable. The movable set of plates is attached to a bellows which acts as an air seal and a low inductance conductor. Atmospheric  pressure pushes the plates fully enmeshed, and the bellows is fully extended. Turning a screw (the shaft) pulls the bellows against atmospheric pressure. This backs out the movable plates from the stationary plates.  It can take 20 turns or so, of the shaft, to go from one extreme to the other, making tuning easy. The shaft needs an insulated shaft coupler.

 

For Litz wire, “3000/46” means it is 3000 individually insulated strands of 46 gauge wire, bundled, twisted and braided in a way so that each strand will occupy all positions within the bundle over length.  Since RF travels along the outside surface of a strand, having many strands, totaling a large amount of surface area, reduces the resistance of the wire. Also, RF current in a wire is influenced by nearby fields adjacent to the wire. Having each strand occupy all positions evenly within the cable distributes this influence among all strands. 

 

Permeability of ferrite includes two components. One component is responsible for concentrating magnetic flux. This increases the inductance of a coil. The other component is responsible for losses. These losses behave like an added resistance in the coil, reducing the tuned Q. Losses get higher with frequency.  Lossy ferrite isn’t always a bad thing.  Some ferrite mixes are deliberately used at frequencies where it is very lossy. It’s very useful for suppression of unwanted energy. Lossy ferrite cores are also an excellent choice for common-mode chokes and bifilar wound 1:1 baluns. 

 

Most variable capacitors are electrically two capacitors in parallel. One capacitance is between the intermeshed plates of the rotor and stator. The other capacitance is between the stator and the capacitor’s frame.

The Q of a variable capacitor is degraded by the contact resistance between the rotor and frame. The good news is that when the plates are fully unmeshed, this resistance is of little consequence because most of the capacitance is directly between the frame and stator.