(for reason of size, broken into two parts)
About spurious signals other than aliases-I believe, Geoff, you have=
also alluded to signals that can occur because of a 'heterodyne' effect. =
Please permit me to speak to this issue in some detail without (hopefully) =
appearing to be an arrogant ass. I believe the salient features of this ty=
pe of artifact are something I understand pretty well. When the electronic=
s has nonlinear characteristics (fundamental requirement for a mixer), then=
two signals interacting via this nonlinearity will generate sum and differ=
ence frequencies. On a related issue, a single pure (monochromatic) signal=
will, when passing through the system that is nonlinear, produce harmonics=
.. (My PhD research was concerned with ultrasonic harmonic generation-simil=
ar to frequency doubling used to convert an invisible infrared laser (YAG) =
beam into a visible beam after passing through a KDP crystal). The amount=
of harmonic(s) that are generated depends on the power. For a single sign=
al, the amount of harmonic content produced depends on the square of the po=
wer in that signal. For two signals that mix, the sum/difference signals a=
re proportional to the product of their individual powers.
The 'verbosity' of the previous paragraph is to try and convince you t=
hat artifacts of the type that would be part of nonlinearity are not nearly=
so important for the list-serve readers as is the artifact of aliasing typ=
e. And yes, you are absolutely correct-the standard procedure for eliminat=
ing an alias is to hardware filter out the offending signal before it can c=
ause problems. Again, however, I want to try and make the point (which I b=
elieve to be correct)-that the smaller the electronics package, the less li=
kely is the probability of offending pickup signals to cause an alias.
I want to also elaborate on the matter of what is required of the off=
ending signal. It must, as I mentioned before, be a persistent signal, suc=
h as 60-Hz pickup. The coherence time (number of cycles through which the =
offending signal exists) is a very important issue, depending on its freque=
ncy compared to the frequency one studies. These two must be related to on=
e another as a ratio of commensurate integers for the alias to exist.
Another of my comments that needs clarification. You mention, Geoff, t=
he possible ridiculous consequence of software 'integration' (a discrete nu=
merical approximation using software, operating by an iterative process). =
You are correct that what comes out of the 'integral' may be useless. The=
'integrated' signal from the VolksMeter is used by me only as a 'cue'. I=
am interested in studying the 'pure' (raw) signals before, during, and aft=
er an earthquake, particularly teleseismic ones. The helicord display of t=
he integrated signal is a convenient way to improve the SNR for teleseismic=
viewing. When I see something interesting in the integrated signal, I wil=
l then download only the non-integrated record. I have never saved the int=
egrated signals because so much information has been lost from them.
More about the VolksMeter-my encouragement for some of you list-serve r=
eaders try and build something (especially a vertical instrument) using its=
operational principles-this was with the sure knowledge that many of you a=
re much better skilled than myself (and probably every other 'professional'=
that I've met) when it comes to building both mechanical devices and also =
electronics. Larry Cochrane put together the first working prototype a cou=
ple of years ago. Hearing what was needed (both mechanical and electronic)=
on a Saturday, he proceeded to have one both built and operational on the =
web the following Monday! I wrote and asked, "do you routinely 'walk on wa=
ter'; if so I want to come out and watch sometime".
You could build such an instrument 'on a shoestring'. The mechanical p=
arts would be 'a breeze' for many of you and I'll bet that you will be surp=
rised at how low the cost is for the electronics. I want to see what migh=
t develop, because my efforts have 'only scratched the surface' of possibil=
ities.
Randall
(for reason of size, broken into two parts)
About spurious signals =
other
than aliases—I believe, Geoff, you have also alluded to signals=
that can
occur because of a ‘heterodyne’ effect. Please permit me =
to
speak to this issue in some detail without (hopefully) appearing to be an
arrogant ass. I believe the salient features of this type of artifact=
are
something I understand pretty well. When the electronics has nonlinea=
r
characteristics (fundamental requirement for a mixer), then two signals
interacting via this nonlinearity will generate sum and difference
frequencies. On a related issue, a single pure (monochromatic) signal
will, when passing through the system that is nonlinear, produce
harmonics. (My PhD research was concerned with ultrasonic harmonic
generation—similar to frequency doubling used to convert an invisible
infrared laser (YAG) beam into a visible beam after passing through a KDP
crystal). The amount of harmonic(s) that are generated depends =
on
the power. For a single signal, the amount of harmonic content produc=
ed
depends on the square of the power in that signal. For two signals th=
at
mix, the sum/difference signals are proportional to the product of their
individual powers.
The ‘verbosity’ o=
f the
previous paragraph is to try and convince you that artifacts of the type th=
at
would be part of nonlinearity are not nearly so important for the list-serv=
e
readers as is the artifact of aliasing type. And yes, you are absolut=
ely
correct—the standard procedure for eliminating an alias is to hardwar=
e
filter out the offending signal before it can cause problems. Again,
however, I want to try and make the point (which I believe to be
correct)—that the smaller the electronics package, the less likely is=
the
probability of offending pickup signals to cause an alias.
I want to also elaborat=
e on
the matter of what is required of the offending signal. It must, as I=
mentioned
before, be a persistent signal, such as 60-Hz pickup. The coherence t=
ime
(number of cycles through which the offending signal exists) is a very
important issue, depending on its frequency compared to the frequency one
studies. These two must be related to one another as a ratio of
commensurate integers for the alias to exist.
Another of my comments that needs
clarification. You mention, Geoff, the possible ridiculous consequenc=
e of
software ‘integration’ (a discrete numerical approximation usin=
g
software, operating by an iterative process). You are correct that wh=
at
comes out of the ‘integral’ may be useless. The
‘integrated’ signal from the VolksMeter is used by me only as
a ‘cue’. I am interested in studying the
‘pure’ (raw) signals before, during, and after an earthquake,
particularly teleseismic ones. The helicord display of the integrated
signal is a convenient way to improve the SNR for teleseismic viewing. =
;
When I see something interesting in the integrated signal, I will then down=
load
only the non-integrated record. I have never saved the integrated sig=
nals
because so much information has been lost from them.
More about the VolksMeter—my
encouragement for some of you list-serve readers try and build something
(especially a vertical instrument) using its operational principles—t=
his
was with the sure knowledge that many of you are much better skilled than
myself (and probably every other ‘professional’ that I’ve
met) when it comes to building both mechanical devices and also electronics=
..
Larry Cochrane put together the first working prototype a couple of years
ago. Hearing what was needed (both mechanical and electronic) on a
Saturday, he proceeded to have one both built and operational on the web th=
e
following Monday! I wrote and asked, “do you routinely ‘w=
alk
on water’; if so I want to come out and watch sometime”. <=
/o:p>
You could build such an instrument
‘on a shoestring’. The mechanical parts would be ‘a
breeze’ for many of you and I’ll bet that you will be surprised=
at
how low the cost is for the electronics. I want to see what mig=
ht
develop, because my efforts have ‘only scratched the surface’ o=
f
possibilities.
Randall