Hello Jim, VLF,
Thank you for your helpful suggestions!
Es tnx for calculating the FS at the earth antennas QTH! This was what i
didn't know how to do ;-)
Yes, i have a low pass filter attached at the input, a simple RC lowpass
(100kOhm, 47pF).
So, if the effective antenna area is about 18000m^2 and the area above
the ground is just 1000m^2, it seems that this kind of antenna could
work quite fine if i can reduce the ground electrode losses! :-)
It will be interesting if i can increase the DHO38 signal by 6 dB when
doubling the distance between the electrodes!
The background noise of my netbooks soundcard in 3 Hz is -105 dB. To
make the arrangement easier to handle, i choose the USBs 5V supply for
the circuit. Maybe this causes the noise i thought today. Thus, i choose
a 9V block battery today as the power supply. The noise decreased to
-100 dB (with short cut at the two antenna input cables), so not much
more than the soundcard noise. If i cannot see DHO38 now i will decrease
the FFT time, e.g. to 0,3 Hz.
Maybe your calcutated -94 dB level can be confirmed in the further
tests. I hope it is even lower of course since that means that the
effective area is even higher ;-)
Will report here...
73, Stefan/DK7FC
Am 15.07.2010 01:40, schrieb James Moritz:
Dear Stefan, LF Group,
Taking a guess at ERP of DHO38 as 100kW, the FS at 411km distance
would be about 5mV/m (if ERP were only 10kW, FS would be 1.7mV/m, if
it were 1MW, 17mV/m, so 5mV is probably not too many dB wrong). The
EMF induced in a 1-turn, 1m^2 loop would be about 2.5uV. From your
description yesterday, this should give a level on the SpecLab display
of -9dB -(44mV/2.5uV)dB which is -94dB, so below your background noise
level, as you discovered.
The noise voltage of a TLC271 is typically 25nV/sqrtHz - this is a
poor match for a low impedance source such as a single-turn loop, but
even so the signal level should be well above the noise. With 3Hz FFT
resolution, op-amp noise might be 50 or 60nV, which would be about
-127dB on the SpecLab scale according to your calibration. Therefore
the op-amp is probably not the limiting factor here, it is more likely
the sound card input is determining the noise floor. So you need to
increase the gain in front of the sound card by 30dB - 40dB to get a
good SNR.
You could do this by:
-Increasing the number of turns in the loop - e.g. a 30 turn loop like
I used for 9kHz reception, which would increase EMF by nearly 30dB (it
would also increase Z of the loop by a factor of about 900, but at VLF
this would still be small compared to the op-amp Zin)
-Put a step-up transformer between loop and op-amp - a 1:30 ratio
would achieve a similar result to having a 30 turn loop.
-Increase the gain of the op-amp stage. To get 30dB gain at 23.4kHz, a
TLC271 has marginal gain-bandwidth product; an OP-27 would be better
in this respect as well as lower noise.
Or some combination of these things. But remember this will also
increase the level of unwanted signals at the soundcard input - if you
have any nearby broadcast stations, you might need to include low-pass
filtering.
BTW, if my guess of 5mV/m is correct, and the loop EMF is 44mV, the
effective loop area would be about 18000m^2...
Cheers, Jim Moritz
73 de M0BMU
|
