Re: VLF beginner.
- Currently with the recommendations of the group I have switched from Mic input on on USB sound to line. Also denoiser is now off.
Positive results are a lower noise floor, a higher density of sferic/tweak activity, the ability to eliminate the use of denoiser in SL. I did have to initiate the P Nicholson hum filter in L4/5 and it seems to be working flawlessly.
Negative results is a narrow peak centering at 6.2Khz that still needs resolving. This was present while using Mic input but not as narrow and strong.
Overall recommendations thus far has made, in my opinion, a marked improvement in audio quality and sferic/tweak reception.
You can view the spectrum at http://vlf.ssbwx.net to get an idea of what I am talking about.
Thanks to the group!!
--- In VLF_Group@yahoogroups.com, Tom Becker <gtbecker@...> wrote:
> On 04/29/2013 08:46 PM, ssshoema wrote:
> > Cat6, 40 feet, to the loop...
> Transformer(s) on the twisted pair? The least you should have with a
> wired loop is an isolation transformer at the sound card Line input to
> balance the pair and loop. Two transformers (an additional one at the
> loop end of the twisted pair) might improve noise, too.
> If you have insufficient signal for the Line in, maybe you need more
> loop area or turns, or make some gain with the transformers (e.g use a
> 8:2kOhm instead of 1:1). You can also, though, add gain to the ADC
> output in SL's L4/L5 amplifier, but it's better to find or make the gain
> ahead of the ADC. What is the loop?
> Denoiser off. Eliminate the noise source, don't hide it.
> ... peak centering at 6.2Khz...That is pretty severe; it sounds like a high-Q filter. Also, what is
the source of the ~1pps up at 15kHz?
You are receiving spherics, certainly. Your stream is about two seconds
ahead of my Cape Coral and Joel's Sebring feed (via Paul's server), so
putting them on opposite speakers makes pattern comparison pretty easy.
You just need to get that peaking under control. You might try feeding
some convenient low-impedance audio - a radio maybe, or a sweeping
signal generator - instead of the loop to see if the peak is caused by
the loop or downstream from it. Just for grins, try turning SLs filters
off altogether briefly to see what happens. Hum will come up, of
course, but does the 6kHz peak disappear, too? Maybe you have an
inadvertent filter curve working?
- With your recommendation went from Mic to Line input and did a step by step from antenna to soundcard. Checked/cleaned all connections.
Installed 1:1 600 ohm audio transformer.
Peak is gone -- Also SNR is much better with some tweaking of line input level and L4/5 AGC and not trying to drive audio so hard with line input instead gaining with AGC. Sounds much better.
Took my dunce hat off and connected to Cape Coral, Sebring, and my stream at the same time and realized that I was just trying to drive audio harder when spheric activity was low. I have tried to match as closely as possible the same "listening level" that the two Florida stations have. In comparison my noise level is still higher but not by much. Thanks for the tips...they worked.
The 1pps at 15 is something I have been struggling to pin down since I have started. It is not constant, some days it's there, some it's not. I have gone thru my house and shut off everything electric one at a time with no success. I would suspect it is from an outside source.
So as it sits from antenna to speaker, Degauss Loop, 40' CAT6, 1:1 audio transformer, Line In at 25%, 48000, SL Filter with notch on, denoiser off, L4/5 AGC at 25db gain and out. Of course my intent is to squeeze all I can so I will be fiddling with SL. As of today I have narrowed bandpass to help limit the 1pps at 15k.
My next plan is to relocate the loop to a higher, clearer part of my yard. I had heard that wrapping it with foil also reduces noise???
Also in SL on a loud spheric the notch filter seems to drop out and return leaving a "notch trail" after the spheric has occurred, any clues?
Hey, thanks for the instructions, you have been of great assistance.
Steve - AD4TH
--- In VLF_Group@yahoogroups.com, Tom Becker <gtbecker@...> wrote:
> > ... peak centering at 6.2Khz...
> That is pretty severe; it sounds like a high-Q filter. Also, what is
> the source of the ~1pps up at 15kHz?
> You are receiving spherics, certainly. Your stream is about two seconds
> ahead of my Cape Coral and Joel's Sebring feed (via Paul's server), so
> putting them on opposite speakers makes pattern comparison pretty easy.
> You just need to get that peaking under control. You might try feeding
> some convenient low-impedance audio - a radio maybe, or a sweeping
> signal generator - instead of the loop to see if the peak is caused by
> the loop or downstream from it. Just for grins, try turning SLs filters
> off altogether briefly to see what happens. Hum will come up, of
> course, but does the 6kHz peak disappear, too? Maybe you have an
> inadvertent filter curve working?
> ... tweaking of line input level...25dB L4/L5 gain sounds high to me. To see what the line input signal
> ... Line In at 25%  L4/5 AGC at 25db gain...
level looks like, attach the Time Domain Scope to L1 and turn it on. If
you see a flat line you probably should add some front-end gain - either
or both with a step-up transformer and more Line mixer gain. To get an
idea of what the level is, turn on the Signal Generator, pick a
frequency of your choice and a level of -10dB, and connect it via the
SPDT switch to the L1 summer; you should see the generated tone in the
scope with peak-to-peak amplitude of~+/-20000 units (of +/-32767). Turn
off and/or disconnect the signal generator and adjust your line input
gain so that the hum (assuming it dominates) is about the same level,
giving you a headroom of 10dB. Assuming the hum dominates, after it
has been notched you'll be left with a smaller signal that you can make
up with L4/L5 gain as required. I no longer use Spectrum Lab for VLF
but, as I recall, I used only 6dB L4/L5 gain to make up the loss of a
pretty-extreme custom filter I used to minimize TWACS interference here
- but I use a whip on a VLF-3, not a loop; your loop might produce a
> ... Peak is gone...It is reduced, I agree, but it sounds to me like there is still a
resonance that needs damping - and a remaining significant noise level.
> ... wrapping it with foil also reduces noise?Yes, a grounded shield layer around the loop can help reduce noise - but
make certain you leave an electrical gap somewhere in that layer or you
will have produced a shorted turn that will reduce the loop sensitivity
> ... on a loud spheric the notch filter seems to drop out and returnleaving a "notch trail"...
Yes, a common artifact of a digital adaptive filter which continually
learns what to remove to produce the notch(es). You'll hear similar
effects on most of the streams; in severe pops, a reverberating artifact
seems to leave the filter first, preceding the processed spheric. One
solution is a series of analog notch filters - or just a high-pass and a
harmonic-tolerant ear - but the digital solution is otherwise so
effective that we accept the artifacts. Another, ideal but probably
impractical, treatment is to move to some place where there is no
utility power available for many miles and, therefore, no hum filter
needed; Antarctica, maybe.
Your loop, though, probably has some directionality (its plane is
vertical, yes?). You might try rotating it on its vertical mount to
find an attitude that might reduce the hum (and, perhaps, help identify
your 15kHz tick source).
It sounds like you're making progress, Steve.