Test post (ignore)

[Bill Bowden]

"Bill Bowden" <bperryb@bowdenshobbycircuits.info> wrote in message
news:nh89ic$28k2$1@adenine.netfront.net...

My news server is getting intermittent and only posts when it wants to.
Just wanted to see if it is working today.

Test post for Saturday May 21. This probably won't go through since it's
Saturday night.






--- news://freenews.netfront.net/ - complaints: news@netfront.net ---

 

[Phil Hobbs]

On 05/21/2016 07:47 PM, Bill Bowden wrote:

"Bill Bowden" <bperryb@bowdenshobbycircuits.info> wrote in message
news:nh89ic$28k2$1@adenine.netfront.net...
My news server is getting intermittent and only posts when it wants to.
Just wanted to see if it is working today.


Test post for Saturday May 21. This probably won't go through since it's
Saturday night.

Worked here.


Any progress on the Fourier stuff?

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Bill Bowden]

"Phil Hobbs" <pcdhSpamMeSenseless@electrooptical.net> wrote in message
news:5740FAEA.1080205@electrooptical.net...

On 05/21/2016 07:47 PM, Bill Bowden wrote:
"Bill Bowden" <bperryb@bowdenshobbycircuits.info> wrote in message
news:nh89ic$28k2$1@adenine.netfront.net...
My news server is getting intermittent and only posts when it wants to.
Just wanted to see if it is working today.


Test post for Saturday May 21. This probably won't go through since it's
Saturday night.


Worked here.


Any progress on the Fourier stuff?

A little progress, but the math is beyond me. I did do a LTspice simulation
of a triangle wave and was able to measure the amplitude of the third
harmonic at about -18dB. That seemed to agree with other references
suggesting it should be about 1/9. Amazing what simulators can do.







--- news://freenews.netfront.net/ - complaints: news@netfront.net ---

 

[Phil Hobbs]

A little progress, but the math is beyond me. I did do a LTspice simulation
of a triangle wave and was able to measure the amplitude of the third
harmonic at about -18dB. That seemed to agree with other references
suggesting it should be about 1/9. Amazing what simulators can do.

If you put a bunch of sine wave sources in series with the coeficients I derived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run a transient sim for a cycle or two, you can look at the convergence of the partial sums directly. The behaviour of the overshoot is pretty interesting..

You need to be comfortable with calculus to understand the full Fourier theory, though.

Cheers

Phil Hobbs

 

On Tuesday, May 24, 2016 at 5:46:18 AM UTC-7, Phil Hobbs wrote:

A little progress, but the math is beyond me. I did do a LTspice simulation
of a triangle wave and was able to measure the amplitude of the third
harmonic at about -18dB. That seemed to agree with other references
suggesting it should be about 1/9. Amazing what simulators can do.

If you put a bunch of sine wave sources in series with the coeficients I derived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run a transient sim for a cycle or two, you can look at the convergence of the partial sums directly. The behaviour of the overshoot is pretty interesting.

You need to be comfortable with calculus to understand the full Fourier theory, though.

Cheers

Phil Hobbs

Overshoot, as in Gibbs Phenomenon?

Michael

 

[Phil Hobbs]

On 05/24/2016 12:03 PM, mrdarrett@gmail.com wrote:

On Tuesday, May 24, 2016 at 5:46:18 AM UTC-7, Phil Hobbs wrote:
A little progress, but the math is beyond me. I did do a LTspice simulation
of a triangle wave and was able to measure the amplitude of the third
harmonic at about -18dB. That seemed to agree with other references
suggesting it should be about 1/9. Amazing what simulators can do.

If you put a bunch of sine wave sources in series with the coeficients I derived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run a transient sim for a cycle or two, you can look at the convergence of the partial sums directly. The behaviour of the overshoot is pretty interesting.

You need to be comfortable with calculus to understand the full Fourier theory, though.

Cheers

Phil Hobbs


Overshoot, as in Gibbs Phenomenon?

Michael

Right. It's a pretty cool demo. With a bit of SPICE behavioural magic
you can put on different window functions and see how the convergence
changes.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Bob Masta]

On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
<pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm>
That's an 8-part series for an on-line mag that used to be
EE Times.

An equivalent treatment of Fourier theory from Daqarta Help
begins at <http://www.daqarta.com/dw_0h0i.htm>

Enjoy!,


Bob Masta

DAQARTA v9.20
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE 8-channel Signal Generator, DaqMusiq generator
Science with your sound card!

 

[Phil Hobbs]

On 05/25/2016 08:07 AM, Bob Masta wrote:

On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.


Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[John Larkin]

On Sat, 14 May 2016 15:41:14 -0700, "Bill Bowden"
<bperryb@bowdenshobbycircuits.info> wrote:

My news server is getting intermittent and only posts when it wants to. Just
wanted to see if it is working today.

The longest threads start with "test post---ignore"


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[George Herold]

On Wednesday, May 25, 2016 at 10:47:19 AM UTC-4, Phil Hobbs wrote:

On 05/25/2016 08:07 AM, Bob Masta wrote:
On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.

One of my favorite freshman physics labs was this Fourier Synthesizer
mad by Pasco. You would add in all the harmonics, amplitude and phase,
and see the result on a 'scope. Of course you can do that all on a
computer now... .but I'm a man who likes to turn knobs :^)

George H.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Bob Masta]

On Wed, 25 May 2016 10:47:15 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 05/25/2016 08:07 AM, Bob Masta wrote:
On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.

Err, those links *were* about Fourier transforms, not
spectrum analyzers. The point is that you don't need
anything more than high-school trig and algebra, at least if
you are working in the digital domain. (After all, that's
what a computer is using, being pretty lousy at calculus!)


Bob Masta

DAQARTA v9.20
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE 8-channel Signal Generator, DaqMusiq generator
Science with your sound card!

 

[George Herold]

On Thursday, May 26, 2016 at 11:51:45 AM UTC-4, Phil Hobbs wrote:

On 05/26/2016 11:32 AM, George Herold wrote:
On Wednesday, May 25, 2016 at 10:47:19 AM UTC-4, Phil Hobbs wrote:
On 05/25/2016 08:07 AM, Bob Masta wrote:
On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.

One of my favorite freshman physics labs was this Fourier Synthesizer
mad by Pasco. You would add in all the harmonics, amplitude and phase,
and see the result on a 'scope. Of course you can do that all on a
computer now... .but I'm a man who likes to turn knobs :^)


Me too. All the elementary discussions (including mine upthread) seem
to get you to Fourier series for periodic functions, and not as far as
actual Fourier transforms.

FFTs are Fourier series decompositions, not transforms. The FFT gives
you samples of the true continuous-time transform in one special case,
i.e. band limited functions that are periodic in the length of the
transform. You're virtually never in that case in real life.

Grin, well except for all the windowing stuff, I don't see much difference.
(Or is your point about the windowing that you have to do?)

I did buy Bracewell's book, but I only gave it a quick skim... and it waits
on my beside table.

George H.

You can certainly do some additional math to compute the deviation of
the FFT from true samples of the actual transform, which is where a lot
of the FFT's utility comes from, but that's definitely calculus.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Phil Hobbs]

On 05/26/2016 07:50 AM, Bob Masta wrote:

On Wed, 25 May 2016 10:47:15 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 05/25/2016 08:07 AM, Bob Masta wrote:
On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.

Err, those links *were* about Fourier transforms, not
spectrum analyzers.

Sorta.

The point is that you don't need
anything more than high-school trig and algebra, at least if
you are working in the digital domain. (After all, that's
what a computer is using, being pretty lousy at calculus!)

I expect we aren't going to agree about that.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Phil Hobbs]

On 05/26/2016 11:32 AM, George Herold wrote:

On Wednesday, May 25, 2016 at 10:47:19 AM UTC-4, Phil Hobbs wrote:
On 05/25/2016 08:07 AM, Bob Masta wrote:
On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.

One of my favorite freshman physics labs was this Fourier Synthesizer
mad by Pasco. You would add in all the harmonics, amplitude and phase,
and see the result on a 'scope. Of course you can do that all on a
computer now... .but I'm a man who likes to turn knobs :^)

Me too. All the elementary discussions (including mine upthread) seem
to get you to Fourier series for periodic functions, and not as far as
actual Fourier transforms.

FFTs are Fourier series decompositions, not transforms. The FFT gives
you samples of the true continuous-time transform in one special case,
i.e. band limited functions that are periodic in the length of the
transform. You're virtually never in that case in real life.

You can certainly do some additional math to compute the deviation of
the FFT from true samples of the actual transform, which is where a lot
of the FFT's utility comes from, but that's definitely calculus.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Phil Hobbs]

On 05/26/2016 12:22 PM, George Herold wrote:

On Thursday, May 26, 2016 at 11:51:45 AM UTC-4, Phil Hobbs wrote:
On 05/26/2016 11:32 AM, George Herold wrote:
On Wednesday, May 25, 2016 at 10:47:19 AM UTC-4, Phil Hobbs wrote:
On 05/25/2016 08:07 AM, Bob Masta wrote:
On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.

One of my favorite freshman physics labs was this Fourier Synthesizer
mad by Pasco. You would add in all the harmonics, amplitude and phase,
and see the result on a 'scope. Of course you can do that all on a
computer now... .but I'm a man who likes to turn knobs :^)


Me too. All the elementary discussions (including mine upthread) seem
to get you to Fourier series for periodic functions, and not as far as
actual Fourier transforms.

FFTs are Fourier series decompositions, not transforms. The FFT gives
you samples of the true continuous-time transform in one special case,
i.e. band limited functions that are periodic in the length of the
transform. You're virtually never in that case in real life.

Grin, well except for all the windowing stuff, I don't see much difference.
(Or is your point about the windowing that you have to do?)

FFTs applied by people who don't know their limitations are one of the
greatest generators of reasonable-looking wrong answers in all of
science and technology.

Getting the sampling and windowing right, and putting hard bounds on the
truncation error, are super important for good measurements.

I did buy Bracewell's book, but I only gave it a quick skim... and it waits
on my beside table.

Great book. Of course I liked Bracewell himself, so it has good
associations for me.

Cheers

Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[George Herold]

On Thursday, May 26, 2016 at 3:38:06 PM UTC-4, Phil Hobbs wrote:

On 05/26/2016 12:22 PM, George Herold wrote:
On Thursday, May 26, 2016 at 11:51:45 AM UTC-4, Phil Hobbs wrote:
On 05/26/2016 11:32 AM, George Herold wrote:
On Wednesday, May 25, 2016 at 10:47:19 AM UTC-4, Phil Hobbs wrote:
On 05/25/2016 08:07 AM, Bob Masta wrote:
On Tue, 24 May 2016 05:46:14 -0700 (PDT), Phil Hobbs
pcdhobbs@gmail.com> wrote:

A little progress, but the math is beyond me. I did do a LTspice simulatio=
n=20
of a triangle wave and was able to measure the amplitude of the third=20
harmonic at about -18dB. That seemed to agree with other references=20
suggesting it should be about 1/9. Amazing what simulators can do.=20

If you put a bunch of sine wave sources in series with the coeficients I de=
rived in my post (odd harmonics only, amplitude 4/(N pi), phase 0) and run =
a transient sim for a cycle or two, you can look at the convergence of the =
partial sums directly. The behaviour of the overshoot is pretty interesting=
.=20

You need to be comfortable with calculus to understand the full Fourier the=
ory, though.

Maybe, but you might want to check out "Gut - Level Fourier
Transforms - Part 1: Everything You Need, You Got In High
School" at <http://www.daqarta.com/eex01.htm
That's an 8-part series for an on-line mag that used to be
EE Times.

Sure thing, you don't have to be a math whiz to know how a spectrum
analyzer works, or to use one. HP AN 150 is another good read on that
point. (There have been several versions of it, all called AN 150.)

The OP's question was about Fourier transforms, though.

One of my favorite freshman physics labs was this Fourier Synthesizer
mad by Pasco. You would add in all the harmonics, amplitude and phase,
and see the result on a 'scope. Of course you can do that all on a
computer now... .but I'm a man who likes to turn knobs :^)


Me too. All the elementary discussions (including mine upthread) seem
to get you to Fourier series for periodic functions, and not as far as
actual Fourier transforms.

FFTs are Fourier series decompositions, not transforms. The FFT gives
you samples of the true continuous-time transform in one special case,
i.e. band limited functions that are periodic in the length of the
transform. You're virtually never in that case in real life.

Grin, well except for all the windowing stuff, I don't see much difference.
(Or is your point about the windowing that you have to do?)

FFTs applied by people who don't know their limitations are one of the
greatest generators of reasonable-looking wrong answers in all of
science and technology.

Getting the sampling and windowing right, and putting hard bounds on the
truncation error, are super important for good measurements.

Grin, We'd made, (not much me.) a new nmr to do chemical shifts
beating in the FID. (an exponential ringdown, with beats.)
Things looked good, my boss could see the beats, on the 'scope
in the time domain.
The 'scope FFT was crap.
Some hamming, or other window. anyway not a flat top.

DSO's with FFT's are a boon to understanding.
At least that's my experience.

George H.

I did buy Bracewell's book, but I only gave it a quick skim... and it waits
on my beside table.

Great book. Of course I liked Bracewell himself, so it has good
associations for me.

Cheers

Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Bill Bowden]

"Bill Bowden" <bperryb@bowdenshobbycircuits.info> wrote in message
news:nh89ic$28k2$1@adenine.netfront.net...

My news server is getting intermittent and only posts when it wants to.
Just wanted to see if it is working today.

Didn't work yesterday. Wondering if it works today. It's a free news
server, so I can't complain too much.






--- news://freenews.netfront.net/ - complaints: news@netfront.net ---

 

[Phil Hobbs]

Grin,  We'd made, (not much me.) a new nmr to do chemical shifts
beating in the FID.  (an exponential ringdown, with beats.)
Things looked good, my boss could see the beats, on the 'scope
in the time domain.
The 'scope FFT was crap.
Some hamming, or other window. anyway not a flat top.

DSO's with FFT's are a boon to understanding.  
At least that's my experience.

A few of my scopes do FFTs, but I don't find them that helpful due to their being crap, as you say.

A lot of that is from their 8-bit digitizers.

Cheers

Phil Hobbs
 

 

[George Herold]

On Friday, May 27, 2016 at 2:32:12 AM UTC-4, Phil Hobbs wrote:

Grin,  We'd made, (not much me.) a new nmr to do chemical shifts
beating in the FID.  (an exponential ringdown, with beats.)
Things looked good, my boss could see the beats, on the 'scope
in the time domain.
The 'scope FFT was crap.
Some hamming, or other window. anyway not a flat top.

DSO's with FFT's are a boon to understanding.  
At least that's my experience.

A few of my scopes do FFTs, but I don't find them that helpful due to their being crap, as you say.

A lot of that is from their 8-bit digitizers.

Cheers

Phil Hobbs
 

Ahh, well the 'scope FFT was crap because of the wrong window.
(Exponential decay was set-up to use the whole scope display,
with it starting at one edge... the hamming window diminished
that part of the signal that had the most information.)
With a Flat top it was OK. (Of course still not as good as
a real spectrum analyzer.)

Hey have you every tried "my" trick to improving the 'scope FFT.
This mainly works on noisy signals.. or even where the signal is all
noise. You trigger (normal) right up at the tippy top of the signal,
(perhaps triggering only on ~10-20% of the possible traces.)
You then push the average button, and your get something that is
very much like the auto-correlation function of the signal.

The FFT of that is much cleaner than a single shot FFT.
(The down side is the high frequency stuff gets attenuated more...
and if you are using this to measure the 3dB point of a filter, then
the 3dB point will be down 6dB on the 'scope FFT display.)

The other cool thing with a 'scope is that some will display both the
time domain and frequency domain at the same time.

George H.

 

[Bill Bowden]

"Bill Bowden" <bperryb@bowdenshobbycircuits.info> wrote in message
news:ni7tej$eql$1@adenine.netfront.net...

"Bill Bowden" <bperryb@bowdenshobbycircuits.info> wrote in message
news:nh89ic$28k2$1@adenine.netfront.net...
My news server is getting intermittent and only posts when it wants to.
Just wanted to see if it is working today.


Didn't work yesterday. Wondering if it works today. It's a free news
server, so I can't complain too much.

Well it did work for a short time yesterday afternoon and then took a dive
later. This test is at 4:30PM PDT.






--- news://freenews.netfront.net/ - complaints: news@netfront.net ---