Re: VLF: E-field receivers in Todmorden

[Markus Vester <markusvester@aol.com>]

There may be another pitfall with the Alpha's: Because three transmitters are sending dashes on the same frequency, the coherent sum of any single spectral line depends on the location of the receiver (that's the essence of a navigation system). At my location, the central line on 11904.7619 Hz happens to have a minimum due to near-cancellation between Krasnodar and Novosibirsk, which also depends on time of day. The superposition may create a noticeable signal difference if the two RX sites are spaced more than a few hundred meters apart.

In principle, calculating MSK station levels by summing power across say 300 Hz should provide a good and accurate reference with low standard deviation. The difficulty may rather be that they are all transmitting above 16 kHz, which might not be perfectly representative of preamp noise at lower VLF.  

Not sure if paralleling FETs is a useful idea to improve preamp performance in an active probe. Paralleling does reduce noise voltage by sqrt(N), but increases gate noise current by the same factor, leading to the same noise figure at N-times reduced source impedance. Especially with a small capacitance probe with many megohms at ELF, paralleling may well worsen rather than improve the noise mismatch. It will be difficult or impossible to see ZEVS and Schumann resonances on a miniwhip, but on it is usually easy to simply connect a larger wire antenna. 

Best 73,
Markus 



-----Ursprüngliche Mitteilung-----
Von: Paul Nicholson <vlf0403@abelian.org>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: So, 10. Sept 2017 18:17
Betreff: Re: VLF: E-field receivers in Todmorden

The old rx is down at the moment, being measured on the
work bench. The system noise floor is about 140nV/m in 1Hz.
That's very high but quite low enough for natural radio (for
which it is mainly used). The natural floor is about 10uV/m.

But after sferic blanking (necessary for weak amateur signals)
the natural floor may be down to 1uV/m and at that level the
system noise from the old rx becomes noticeable. One reason
for the new rx, which has about 5 to 10nV system noise depending
on the front-end bias resistance I eventually decide to use.

Unfortunately only the Alpha signals from the North East
provide a suitable reference. Other signals eg MSK are too
wide and amateur signals are too weak.

Bernd, DF9RB wrote:

> I plan to build this amplifier:
>
http://dg4rbf.lima-city.de/Ultra%20Low%20Noise%20AMP%20V1.7f%20_WebSeite.pdf

> Please your opinion about this concept?

Merging the currents from multiple front-end devices is a
conventional way to achieve ultra low noise and this looks
like a good implementation.

I tried this technique with two devices but the doubling of the
input capacitance reduced the signal by as much as the noise
was lowered, no overall benefit (in fact slightly worse).
This kind of thing shows up when you refer the system noise
to field strength V/m instead of just input voltage.

The system noise voltage is far lower than necessary and the
gain is very high. Perhaps it is intended to work with a very
small probe antenna. In that case the input capacitance is
likely to cause a significant division of the signal voltage.

I wonder how it performs in other respects. Dynamic range and
intermod.

Ultimately, a VLF pre-amp must be designed for a particular
antenna. Measure the noise when the input is loaded by
the equivalent antenna capacitance, and measure gain with
a low impedance test signal injected through that capacitance.

PS, this afternoon the old rx is back into operation.
--
Paul Nicholson
--