I've been thinking about this reply, trying to understand it, and now I think I
finally do. The interference was a single frequency, but it was a square wave,
and square waves can be created with a series of different frequencies of sine
waves, hence the regular bands shown on the spectrogram, which I assume thinks
everything is constructed from sine waves. Correct?
I generated a square wave in Audacity, in a blank track, and sure enough the
spectrogram shows many bands. Some 30 year old synapse connections, created in
a physics class and never used again, have finally made themselves useful.
But I don't know what 1/3 octave spectrum analysis is. Is it something I can do
with Audacity or some other easily available program? And what will it show me?
Peter Shute
From: On
Behalf Of Eric Benjamin
Sent: Monday, 12 March 2012 10:00 AM
To:
Subject: Re: [Nature Recordists] Re: Initial SASS test shows bands of harmonics
> I've just proven it's interference from the M10
I've seen this sort of noise from LCD displays before. Even acoustic noise
generated by the display strobing circuit in some cases.
I subjected the file to some additional analysis and there are alternating
positive and negative impulses at a period of 209 samples. At 44100 Hz sample
rate, that implies a strobing frequency of 211 Hz. The harmonics show up with a
spacing of approximately 400 Hz because the original waveform was a square wave
and square waves have only odd harmonics.
The interfering noise isn't particularly audible because it's masked at lower
frequencies by the ambient noise, and because at higher frequencies many
harmonics fall within a single critical band, making it difficult for the ear to
resolve the frequency of the noise. But it is audible as an increased noise
level in the right channel. This is one of those cases where 1/3 octave
spectrum analysis gives a much better picture of what's going on than FFT
analysis.
> perhaps <snip> it's caused by the PIP struggling with 4 capsules
It seems unlikely. If that were the case the noise would occur regardless of
where the recorder was positioned. And note that 4 capsules won't draw 4X as
much current as only 1 capsule. If the voltage on the capsules were the same it
would draw 4X current, but since the power in PIP is typically supplied through
a 2.2 kOhm resistor, the voltage across the capsule will drop and the operating
point of the FET within the capsule will change. In a typical scenario with 3 V
battery supply and 2.2 kOhm resistor, the most current that could be drawn would
be 1.4 mA, but that would imply a short circuit and thus zero voltage across the
capsules.
> Is it safe to just put a voltmeter across a capsule to check it?
No problem.
Hope this helps!
Eric
----- Original Message ----
From: Peter Shute <<pshute%40nuw.org.au>>
To:
"<naturerecordists%40yahoogroups.com>"
<<naturerecordists%40yahoogroups.com>>
Sent: Sun, March 11, 2012 12:09:36 PM
Subject: RE: [Nature Recordists] Re: Initial SASS test shows bands of harmonics
The clipped noise is from me handling the array as I wrap the towel over it.
I've just proven it's interference from the M10. With the recorder on top of the
array , the effect is reproduced. With the recorder upside down it's even
stronger. With the recorder inside the array, pressed again the wiring between
left and right under the top, it's very strong, and I can clearly hear the hum
from it.
Interestingly, with it insdie, pressed against the cable but upside down, the
effect is nowhere near as strong. Thus, the interference is coming from
underneath the M10.
Even 1m away, I can see the regular spikes on a frequency analysis, but it's too
weak to see on the spectrogram.
Just to be sure it's electrical interference, not acoustic, I held it in front
of the mics, 1cm away, and the effect is weak. Therefore it's electrical
interference from underneath the M10.
The effect is on both channels, but is far stronger in the right channel, I
assume because I ran the wireing from the right over to the left before joining
to the cable I plug into the recorder, and so there are a few more inches of
wiring on that channel.
But perhaps Raimund is right, that it's caused by the PIP struggling with 4
capsules. Can anyone recommend a battery circuit I could try? I was going to
look into this anyway. Is it safe to just put a voltmeter across a capsule to
check it?
Now to get out of this house, and try some more tests at sunrise.
Peter Shute
From:
<naturerecordists%40yahoogroups.com>
<naturerecordists%40yahoogroups.com>] On
Behalf Of Raimund
<raimundspecht%40googlemail.com>]
Sent: Monday, 12 March 2012 4:46 AM
To: <naturerecordists%40yahoogroups.com>
Subject: [Nature Recordists] Re: Initial SASS test shows bands of harmonics
Because the noise obviously starts on the first occurence of the broad-band and
clipped noise (which does not seem to be an acoustic signal), I would guess that
it might have something to do with PIP power provided by the recorder. It could
be that the PIP supply circuitry just does not work properly due to the heavy
load by the four capsules. So, you might try to power the capsules by an
external battery.
Raimund
--- In
<naturerecordists%40yahoogroups.com><naturerecordists%40yahoogroups.com>,
Peter Shute <> wrote:
>
> The link to the sonogram image works now. The album wasn't set for public
>viewing, I assumed that setting was the default. Sorry about that.
>
> https://picasaweb.google.com/pshute2/March112012#5718478097647090786
------------------------------------
"While a picture is worth a thousand words, a
sound is worth a thousand pictures." R. Murray Schafer via Bernie Krause.
Yahoo! Groups Links
|