For this particular loop I only needed a fixed spot on 30 metres (which is so narrow it needs no retuning). But now that you mention it for coverage of all of HF it would be an interesting exercise to design a tunable one.
For the large passive loop I am fiddling with at the moment it would be quite easy to add varactor tuning. Just add a capacitor to block the DC at the cold-side of transformer secondary (using the secondary itself as an RFC), then a high-value resistor from that RF bypassed point to feed the DC bias to the common cathodes of a pair of head-to-head varactors across the loop. That obviously references one side of the loop resonator to the coax braid which may distort the pattern a bit unless the loop is completely electrostatically shielded or you use another high-value resistor to return the DC. An alternative might be to reflect the capacitance to the loop inductance through another transformer to help maintain balance. Not sure how much that would compromise the loop Q or if the unbalance is even that significant to worry about in the first place?
The hardest part for an active loop would be controlling it over just the coax while still powering the buffer. Extra wiring makes it trivial, but is cheating :). One kinda ugly solution would be a regulator in the loop pre-amp set at some minimum supply voltage, say 5-8 volts, then using the difference between that rail and the current supply voltage to bias the varactors. Production of the variable and fairly high voltage (say 24-36 volts) should be easy enough, not much current would be required so the large drop across components at the highest voltage end of the tuning range should not cause excessive dissipation. Power efficiency of such an approach would be terrible, but total power is low.
You would probably want to avoid any more efficient approaches (say switch-mode supplies or digital signalling) which may cause excessive noise in the signals, especially if placed in close proximity to the high-impedance loop buffer. You could just deal with the change of buffer properties associated with varying Vds/Vce at a constant bias current. The Vds limit of the J310 is 25 volts, if you want a minimum Vds of ~3 volts that gives you about 20 volts of tuning range given typical Vpp. The device's gain will be quite different at each end of the tuning range unfortunately, and probably not the way you'd want it (there will be less gain at lower frequencies where the loop is smaller wrt wavelength). You can more closely approach 0 bias across the varactors by lifting the cold side above ground towards the minimum supply voltage. Q degrades with strong signals the smaller the bias you have but I don't think that would be a major problem for typical signals. If you are going to have such strong signals that parametric intermodulation distortion via the varactors is a problem then the amplifier may be in trouble anyway.
Along these lines the current loop amplifier might need protection from adjacent transmitters. The RF voltage across its high-Q circuit could be very large in a strong near-field from a TX antenna, easily frying the JFETs (25 volts Vgs limit). If mounted near your TX antenna you'd probably want a pair of 1N4148s across the loop to protect the amplifier.
23rd April 2010 23:24
Alan Yates wrote...
Peter,
For this particular loop I only needed a fixed spot on 30 metres (which is so narrow it needs no retuning). But now that you mention it for coverage of all of HF it would be an interesting exercise to design a tunable one.
For the large passive loop I am fiddling with at the moment it would be quite easy to add varactor tuning. Just add a capacitor to block the DC at the cold-side of transformer secondary (using the secondary itself as an RFC), then a high-value resistor from that RF bypassed point to feed the DC bias to the common cathodes of a pair of head-to-head varactors across the loop. That obviously references one side of the loop resonator to the coax braid which may distort the pattern a bit unless the loop is completely electrostatically shielded or you use another high-value resistor to return the DC. An alternative might be to reflect the capacitance to the loop inductance through another transformer to help maintain balance. Not sure how much that would compromise the loop Q or if the unbalance is even that significant to worry about in the first place?
The hardest part for an active loop would be controlling it over just the coax while still powering the buffer. Extra wiring makes it trivial, but is cheating :). One kinda ugly solution would be a regulator in the loop pre-amp set at some minimum supply voltage, say 5-8 volts, then using the difference between that rail and the current supply voltage to bias the varactors. Production of the variable and fairly high voltage (say 24-36 volts) should be easy enough, not much current would be required so the large drop across components at the highest voltage end of the tuning range should not cause excessive dissipation. Power efficiency of such an approach would be terrible, but total power is low.
You would probably want to avoid any more efficient approaches (say switch-mode supplies or digital signalling) which may cause excessive noise in the signals, especially if placed in close proximity to the high-impedance loop buffer. You could just deal with the change of buffer properties associated with varying Vds/Vce at a constant bias current. The Vds limit of the J310 is 25 volts, if you want a minimum Vds of ~3 volts that gives you about 20 volts of tuning range given typical Vpp. The device's gain will be quite different at each end of the tuning range unfortunately, and probably not the way you'd want it (there will be less gain at lower frequencies where the loop is smaller wrt wavelength). You can more closely approach 0 bias across the varactors by lifting the cold side above ground towards the minimum supply voltage. Q degrades with strong signals the smaller the bias you have but I don't think that would be a major problem for typical signals. If you are going to have such strong signals that parametric intermodulation distortion via the varactors is a problem then the amplifier may be in trouble anyway.
Along these lines the current loop amplifier might need protection from adjacent transmitters. The RF voltage across its high-Q circuit could be very large in a strong near-field from a TX antenna, easily frying the JFETs (25 volts Vgs limit). If mounted near your TX antenna you'd probably want a pair of 1N4148s across the loop to protect the amplifier.
Regards,
Alan