The completed choke measures 8.9 uh. 8.9 uh = 2803 ohms of XL @ 50.125 mhz.
Did you take your 8.9 uH measurement at 50 MHz?
I ask because you initially said that the impedance magnitude at 50 MHz was 125 ohms. Now you are saying that the inductive reactance is 2803 ohms. If the first number is for a single bead, then the second number would be for more than 30 beads. Something doesn't jive.
- Glenn W9IQ
If your 125 ohms impedance magnitude at 50 MHz is accurate for a single bead of type 43 material, then the complex impedance will be about 116+j46 for a single bead. You can see will have much more resistance than reactance at that frequency.
Five beads will then provide about 230 ohms of reactance for an equivalent inductance of ~0.73 uH at 50 MHz.
- Glenn W9IQ
You are correct of course. Measurement taken at just 1 khz. But the 125 ohm Z @ 50 ohms is accurate. Their graph depicts aprx 60 ohms of XL per bead @ 50 mhz.... or 300 ohms for 5 of em.
One last thought on this topic. The bead structures from Fair-Rite are not tightly controlled as they are intended for general suppression. You can experience a ±25% XL variation. If your specific XL requirement is tight, you may need to plan an extra bead as insurance.
- Glenn W9IQ
Excellent point. Typ I heard +/- 20% variation. N3RR bought 700 of those type 31 (2.4" OD) toroids....and the tolerance was all over the map, and his all came from the same lot number.
On this 6M amp, it will work. The way the beads are configured, there was no way to get more than 5 of em on there.
If the supercon connectors were shifted more to the right, more cores could have been installed. But that would also add more stray C from beads to the chassis...+ a 90 deg bend in the corner. As is, stray C is 50 pf...added to the 42 pf input C = 92 pf.....which has to be subtracted from the target value of 127 pf for the C2 tune cap.
On any 160-10m amp, 8-15 cores would be optimum...if they will fit. The increase in C would not be an issue...except for 10m, where 225 pf is the target value for C2. I finally figured out a way to space the cores away from the chassis, and still keep them in place. That would reduce the stray C to a low value. It's doable.
Paralleling 40 amp fil chokes using 8 ga magnet wire, (bifilar wound on type 43, 8" x 1/2" rods) is more labour intensive. I initially wind them on 1/2" OD X 8" long alum tubes. The type 43 rods are too brittle to wind on. The trick here is to parallel any one winding from rod 'A' with any one winding from rod 'B'... and vice versa. That ensures bifilar current flow.
The 30 amp chokes, wound with 10 ga wire, measured 38 uh (1 khz) on each winding. The 40 amp chokes, wound with 8 ga wire, measured 30 uh per winding. Interesting enough, when top pair of wires of just one choke are shorted...and ditto with the bottom pair of wires, it's still 38 uh. The coefficient of coupling is 100%. However, when the pair of chokes are paralleled as described in above paragraph, the measured uh drops to 30 uh on the 30 amp chokes...(and 22 uh on the 40 amp chokes).
This is because the rods are spaced a bit apart, like 1.5", inside to inside...and coefficient of coupling is far less. I had used just a single pair of standoffs, and 4 x plastic cable clamps, to mount 2 x rods to one pair of standoffs. The better method would be to use 4 x standoffs, and put zero space between the rods, the the coupling would increase...and uh would increase a bunch.
If the rods are spaced 1' apart, like on the wooden bench, and paralleled with 12" wires, then the inductance, does indeed, drop in half.....since the co-eff of coupling is then non existent.
