On Thu, 29 Nov 2012 14:43:33 -0500, rickman wrote:

On 11/27/2012 11:01 PM, Arid ace wrote:
On Mon, 26 Nov 2012 16:25:37 -0500, wrote:

Just to give an update...

[...]

I did the calculations and at 100 uV/m with 50 foot of coax in a two
turn loop I get a bit over 1 uV, not so good. I'm going to see what I
can find about the signal level from ferrite cores. I had the
impression they would be lower, but I seem to recall better numbers than
this in the ones I've seen before.

Rick

When I was working with loop antennas, it became rather obvious that
optimal receiving loops too are 1 turn, with material of the largest practical
circumference. That one turn loop was coupled to one turn as the primary of a
coil / tuned transformer and on VLF / LF that concerned a potcore of the proper
material. The amount of turns for the secondary could be as much as 100 and
with a Q factor of 100 or more that meant at least 10,000 times the voltage of
the one turn loop - more than enough.

Jan

I have come to a similar conclusion, but not about using a single loop.
The single loop is the best antenna if your constraint is the total
length of cable. But if your limitation is the size of the loop, it
will be improved by more turns at the same diameter.

I used Al strip, bent in a circle, for the loop (used from 0.1 to 30 MHz). It
was cheap and easy but making a multi-turn loop would have been quite a task.

When I wrote the message you replied to I had left the Q factor out of
the equation. I calculated the Q factor at slightly over 100
considering just the resistance of the center conductor. I am also
aware of the idea of coupling using a ferrite core. I don't know
anything about ferrite cores (I suppose I'll need to learn that soon)
and I don't know what the optimal turns ratio would be. I assume that
the resistance of the load reflects back into the antenna as a parallel
resistance? Using your numbers that would mean a 1 Mohm input impedance
would result in a 10 kohm load. At the frequency of 60 kHz and with an
antenna inductance of about 90 uH that gives a reactance of 34 ohms. So
10 kohms shouldn't be a problem. But I read one reference that
indicated the Q needs to be factored in as well which means it would be
100 ohms vs. 33 ohms, and so 1 Mohm might not be high enough.

The antique booklet I have on potcores (Soft Ferrites, October 1973) lists the Q
realized with a core of certain dimensions as a function of ue of the material.
At 60 KHz, a P 11/7 core with material FXC 3B7/3H1 and ue=68 has the lowest Q,
220. The use of material with ue=220 delivers a Q of 420. This might serve as an
estimate of what could be expected: the 11/7 is the smallest potcore and the
bigger ones result in higher maximum Q, to well over 1,000.

When using a FET device as input, at 60 KHz the input impedance can be
neglected.

The input impedance of the device I am using is not characterized. I'll
have to measure it, but I expect it will be rather high, easily 10 Mohm
which should be good enough (30x).

The impedance of a source-follower will be much higher than 10 M - at 60 KHz.

I'm also not sure if there will be a loss of Q due to capacitance of the
cable. I haven't found a clear reference on this. Some say
inter-winding capacitance will reduce the Q, but I don't think this will
be subject to inter-winding capacitance because of the shield. So the
question is will the distributed capacitance to ground create any sort
of problem? This cable has 16.2 pF per foot and I plan to use 50 feet.
I expect this will simply appear as part of the resonating capacitance.

Rick

With a one turn loop coupled to the one turn primary of an RF transformer,
there's no need for a shield. It's fairly easy to provide the piece of wire used
for the primary with a center tap that will be grounded so the loop is
symmetrical. That also obviates the need for a differential amplifier. A 50 feet
loop is huge. I once (in Europe) designed a frequency standard synchronized to
DCF77 and reception was OK at distances over 2,000 Km with just a ferrite rod of
12 cm length and 1 cm diameter. In order to keep it easy for constructors, the
coil had just 140 turns with 2n2 to resonate the LC. I don't think the approach
for WWVB at 60 KHz could be much different (apart from the modulation format).
I had designed the ferrite antenna amp as a small device that could be fed via
coax, so one could install it in a place with the lowest noise.

Jan