Ask a question - edaboard.com
elektroda.net NewsGroups Forum Index - Electronics Design - DDS Selection
Goto page Previous 1, 2, 3, 4, 5, 6, 7, 8 Next
Jamie
Guest
Guest
Wed Jan 25, 2012 1:04 am
John Larkin wrote:
Quote:
On Tue, 24 Jan 2012 15:00:41 -0800 (PST), Bill Sloman
bill.sloman_at_ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
Bill Sloman wrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
The real difference between the stuff I design, and the stuff you used
to design, is that most of my stuff sells. And, of course, I'm still
designing.
John
Hopefully with new applications and not fixing exiting ones!
bill.sloman_at_ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
Bill Sloman wrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
The real difference between the stuff I design, and the stuff you used
to design, is that most of my stuff sells. And, of course, I'm still
designing.
John
Hopefully with new applications and not fixing exiting ones!
Jamie
George Herold
Guest
Guest
Wed Jan 25, 2012 1:45 am
On Jan 24, 1:40 pm, Mikko OH2HVJ <oh2...@sral.fi> wrote:
Quote:
George Herold <gher...@teachspin.com> writes:
So how 'good' can you do Mikko? I've looked at these really tiny
changes in our optical pumping apparatus,
(a straight DC light level measurement, no modulation or lockin) and
seen changes at smaller than a part in 10^5. (That was averaging
several sweeps though.. Signal to noise in one sweep is about 1 at
changes of a part in 10^5.)
The first setup we built about 2 years ago got 0.00001 AU peak-to-peak noise
over 10 minute period with 1s sampling, about 400Hz simple LED source.
Our first try was a DC measurement and 1:10^5 sounds very good for DC!
So how 'good' can you do Mikko? I've looked at these really tiny
changes in our optical pumping apparatus,
(a straight DC light level measurement, no modulation or lockin) and
seen changes at smaller than a part in 10^5. (That was averaging
several sweeps though.. Signal to noise in one sweep is about 1 at
changes of a part in 10^5.)
The first setup we built about 2 years ago got 0.00001 AU peak-to-peak noise
over 10 minute period with 1s sampling, about 400Hz simple LED source.
Our first try was a DC measurement and 1:10^5 sounds very good for DC!
Well there is a 'bit of DC drift.
I've got pictures! Source is a Rubidium lamp, through a Rb reference
cell*, a little less than 10uA of photocurrent. DC level is near 10V,
a gain of 1000 and a 100ms lowpass filter. This is a sweep through
zero magnetic field.
http://imageshack.us/photo/my-images/191/tek0009.png/
There's a little W shaped thing....About 50mV of signal in 50mV of
noise. The drift is due to the temperature drift of the Rb cell.
Here's a similar signal AC coupled and averaged.
http://imageshack.us/photo/my-images/593/tek0024.png/
I 'claim' the broad dip is an optical pumping absorption and the goes
uppy part in the center is CPT. Weird quantum stuff done without a
laser.
George H.
*there's also an interference filter (picks out one spectral line) and
a linear polarizer before the cell... optics details...
Quote:
Now we're doing a digital lock-in, but that is more due to low light available.
We get 0.0002 AU peak-to-peak noise over 100s with 800Hz LED, 200ms measurements.
--
Mikko
spflanze
Guest
Guest
Wed Jan 25, 2012 1:46 am
On Jan 23, 3:05 pm, spflanze <art...@wavenet.org> wrote:
Quote:
On Jan 23, 11:59 am, Mikko OH2HVJ <oh2...@sral.fi> wrote:
spflanze <art...@wavenet.org> writes:
We are looking for the smallest possible changes in transmissivity. So
I am looking to do state of the art in noise performance.
What is your target on transmissivity change and available light power ?
Can you take part of the light before the chemistry and use that as
the reference for the intensity ? This with lock-in detection (done
in HW or SW) has given good results for us in similar measurements.
--
Mikko OH2HVJ
The photodiode would look at the LED sideways, which would be outside
its lens's cone, but would still have plenty of light due to its
proximity. After Bill Sloman's response I am not sure about noise
suppression. I would expect a photodiode in such an arrangement to get
more light from some areas of the light emitting surface than others.
I have some lab experience where the feedback seemed to suppress noise
on the output of a Luxeon III. I do not know where the monitoring
photodiode was relative to this LED because both were in a sealed
unit. The two LEDs being considered now are:
http://partfinder.avagotech.com/Avago/Avago.jsp#!partSearch=&partno=a....http://www.cree.com/products/pdf/LEDlamps/C503B%281079%29.pdf
There is not yet a specific target.
spflanze <art...@wavenet.org> writes:
We are looking for the smallest possible changes in transmissivity. So
I am looking to do state of the art in noise performance.
What is your target on transmissivity change and available light power ?
Can you take part of the light before the chemistry and use that as
the reference for the intensity ? This with lock-in detection (done
in HW or SW) has given good results for us in similar measurements.
--
Mikko OH2HVJ
The photodiode would look at the LED sideways, which would be outside
its lens's cone, but would still have plenty of light due to its
proximity. After Bill Sloman's response I am not sure about noise
suppression. I would expect a photodiode in such an arrangement to get
more light from some areas of the light emitting surface than others.
I have some lab experience where the feedback seemed to suppress noise
on the output of a Luxeon III. I do not know where the monitoring
photodiode was relative to this LED because both were in a sealed
unit. The two LEDs being considered now are:
http://partfinder.avagotech.com/Avago/Avago.jsp#!partSearch=&partno=a....http://www.cree.com/products/pdf/LEDlamps/C503B%281079%29.pdf
There is not yet a specific target.
I am going to build so the light coming out the sides is in a small
cavity with white walls for an integrating sphere effect. The
photodiode would be at the cavity wall. I expect this would make the
monitoring photodiode's current more representative of emissions from
the entire emitting surface. The size of the cavity is going to be
limited by available space.
Another idea I will look into is to have a surface mount led mounted
over a hole in the PCB, and on the other side of the hole have a
surface mount monitoring photodiode. What I am not sure about is how
much light comes out the underside of a surface mount LED. This also
means there will be no lens in the LED package, and I am not sure how
accepting the optics designer will be of that.
George Herold
Guest
Guest
Wed Jan 25, 2012 1:50 am
On Jan 24, 2:06 pm, Bill Sloman <bill.slo...@ieee.org> wrote:
Quote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
--
Bill Sloman, Nijmegen- Hide quoted text -
- Show quoted text -
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
--
Bill Sloman, Nijmegen- Hide quoted text -
- Show quoted text -
Hmm, in a control application you may want the wider bandwidth.
I've got a piezo + (spring and mass) that sings around 1kHz....
What happens if I notch it out of the control loop?
George H.
George Herold
Guest
Guest
Wed Jan 25, 2012 2:48 am
On Jan 24, 6:46 pm, spflanze <art...@wavenet.org> wrote:
Quote:
On Jan 23, 3:05 pm, spflanze <art...@wavenet.org> wrote:
On Jan 23, 11:59 am, Mikko OH2HVJ <oh2...@sral.fi> wrote:
spflanze <art...@wavenet.org> writes:
We are looking for the smallest possible changes in transmissivity. So
I am looking to do state of the art in noise performance.
What is your target on transmissivity change and available light power ?
Can you take part of the light before the chemistry and use that as
the reference for the intensity ? This with lock-in detection (done
in HW or SW) has given good results for us in similar measurements.
--
Mikko OH2HVJ
The photodiode would look at the LED sideways, which would be outside
its lens's cone, but would still have plenty of light due to its
proximity. After Bill Sloman's response I am not sure about noise
suppression. I would expect a photodiode in such an arrangement to get
more light from some areas of the light emitting surface than others.
I have some lab experience where the feedback seemed to suppress noise
on the output of a Luxeon III. I do not know where the monitoring
photodiode was relative to this LED because both were in a sealed
unit. The two LEDs being considered now are:
http://partfinder.avagotech.com/Avago/Avago.jsp#!partSearch=&partno=a...
There is not yet a specific target.
I am going to build so the light coming out the sides is in a small
cavity with white walls for an integrating sphere effect. The
photodiode would be at the cavity wall. I expect this would make the
monitoring photodiode's current more representative of emissions from
the entire emitting surface. The size of the cavity is going to be
limited by available space.
Another idea I will look into is to have a surface mount led mounted
over a hole in the PCB, and on the other side of the hole have a
surface mount monitoring photodiode. What I am not sure about is how
much light comes out the underside of a surface mount LED. This also
means there will be no lens in the LED package, and I am not sure how
accepting the optics designer will be of that.- Hide quoted text -
- Show quoted text -
On Jan 23, 11:59 am, Mikko OH2HVJ <oh2...@sral.fi> wrote:
spflanze <art...@wavenet.org> writes:
We are looking for the smallest possible changes in transmissivity. So
I am looking to do state of the art in noise performance.
What is your target on transmissivity change and available light power ?
Can you take part of the light before the chemistry and use that as
the reference for the intensity ? This with lock-in detection (done
in HW or SW) has given good results for us in similar measurements.
--
Mikko OH2HVJ
The photodiode would look at the LED sideways, which would be outside
its lens's cone, but would still have plenty of light due to its
proximity. After Bill Sloman's response I am not sure about noise
suppression. I would expect a photodiode in such an arrangement to get
more light from some areas of the light emitting surface than others.
I have some lab experience where the feedback seemed to suppress noise
on the output of a Luxeon III. I do not know where the monitoring
photodiode was relative to this LED because both were in a sealed
unit. The two LEDs being considered now are:
http://partfinder.avagotech.com/Avago/Avago.jsp#!partSearch=&partno=a...
There is not yet a specific target.
I am going to build so the light coming out the sides is in a small
cavity with white walls for an integrating sphere effect. The
photodiode would be at the cavity wall. I expect this would make the
monitoring photodiode's current more representative of emissions from
the entire emitting surface. The size of the cavity is going to be
limited by available space.
Another idea I will look into is to have a surface mount led mounted
over a hole in the PCB, and on the other side of the hole have a
surface mount monitoring photodiode. What I am not sure about is how
much light comes out the underside of a surface mount LED. This also
means there will be no lens in the LED package, and I am not sure how
accepting the optics designer will be of that.- Hide quoted text -
- Show quoted text -
Sounds like you need to cobble together some prototypes.
George H.
John Larkin
Guest
Guest
Wed Jan 25, 2012 3:08 am
On Tue, 24 Jan 2012 19:04:57 -0500, Jamie
<jamie_ka1lpa_not_valid_after_ka1lpa__at_charter.net> wrote:
Quote:
John Larkin wrote:
On Tue, 24 Jan 2012 15:00:41 -0800 (PST), Bill Sloman
bill.sloman_at_ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
Bill Sloman wrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
The real difference between the stuff I design, and the stuff you used
to design, is that most of my stuff sells. And, of course, I'm still
designing.
John
Hopefully with new applications and not fixing exiting ones! :)
Jamie
On Tue, 24 Jan 2012 15:00:41 -0800 (PST), Bill Sloman
bill.sloman_at_ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
Bill Sloman wrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
The real difference between the stuff I design, and the stuff you used
to design, is that most of my stuff sells. And, of course, I'm still
designing.
John
Hopefully with new applications and not fixing exiting ones! :)
Jamie
Well, I spent a bunch of today changing some code that I wrote in
1996. Some 20-bit serial DACs went obsolete, so we replaced them with
a 16-bit DAC and a CPLD to shuffle the bits. But there's a glitch, and
the easiest fix is to modify the 68K assembly code that drives the
DACs. Having a product line is like having kids or pets.
John
George Herold
Guest
Guest
Wed Jan 25, 2012 5:38 am
On Jan 24, 9:08 pm, John Larkin
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
Quote:
On Tue, 24 Jan 2012 19:04:57 -0500, Jamie
jamie_ka1lpa_not_valid_after_ka1l...@charter.net> wrote:
John Larkin wrote:
On Tue, 24 Jan 2012 15:00:41 -0800 (PST), Bill Sloman
bill.slo...@ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
Bill Sloman wrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
The real difference between the stuff I design, and the stuff you used
to design, is that most of my stuff sells. And, of course, I'm still
designing.
John
Hopefully with new applications and not fixing exiting ones! :)
Jamie
Well, I spent a bunch of today changing some code that I wrote in
1996. Some 20-bit serial DACs went obsolete, so we replaced them with
a 16-bit DAC and a CPLD to shuffle the bits. But there's a glitch, and
the easiest fix is to modify the 68K assembly code that drives the
DACs. Having a product line is like having kids or pets.
jamie_ka1lpa_not_valid_after_ka1l...@charter.net> wrote:
John Larkin wrote:
On Tue, 24 Jan 2012 15:00:41 -0800 (PST), Bill Sloman
bill.slo...@ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
Bill Sloman wrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
The real difference between the stuff I design, and the stuff you used
to design, is that most of my stuff sells. And, of course, I'm still
designing.
John
Hopefully with new applications and not fixing exiting ones! :)
Jamie
Well, I spent a bunch of today changing some code that I wrote in
1996. Some 20-bit serial DACs went obsolete, so we replaced them with
a 16-bit DAC and a CPLD to shuffle the bits. But there's a glitch, and
the easiest fix is to modify the 68K assembly code that drives the
DACs. Having a product line is like having kids or pets.
I always feel like I'm wheeling out a new baby, when some piece of
equipment, passes testing and goes to shipping.
George H.
Quote:
John- Hide quoted text -
- Show quoted text -
John Larkin
Guest
Guest
Wed Jan 25, 2012 5:54 am
On Tue, 24 Jan 2012 20:08:19 -0800 (PST), spflanze
<artist_at_wavenet.org> wrote:
Quote:
On Jan 24, 10:10 am, spflanze <art...@wavenet.org> wrote:
On Jan 23, 10:37 pm, spflanze <art...@wavenet.org> wrote:
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
I was wrong about the white noise approximation applying to DDS
quantization noise. Because the quantization noise is synchronous with
the sine wave there is no psuedo randomization and so no white noise.
The harmonics will be clearly defined. My decision not to use the DDS
and instead input a square wave into a switched capacitor filter still
stands because I think I will still get a good result and it will be
easy to generate the square wave from the switching frequency I must
generate anyway. I think the DDS will be an additional part with
little benefit.
On Jan 23, 10:37 pm, spflanze <art...@wavenet.org> wrote:
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
I was wrong about the white noise approximation applying to DDS
quantization noise. Because the quantization noise is synchronous with
the sine wave there is no psuedo randomization and so no white noise.
The harmonics will be clearly defined. My decision not to use the DDS
and instead input a square wave into a switched capacitor filter still
stands because I think I will still get a good result and it will be
easy to generate the square wave from the switching frequency I must
generate anyway. I think the DDS will be an additional part with
little benefit.
Switched-cap filters are noisy and nasty. They also alias everything
in sight, including their own power supply noise. And they make big
output spikes.
All you need is a modest analog lowpass filter ahead of the ADC, just
to keep the noise down.
John
spflanze
Guest
Guest
Wed Jan 25, 2012 5:57 am
On Jan 23, 3:31 pm, spflanze <art...@wavenet.org> wrote:
Quote:
On Jan 23, 11:51 am, Bill Sloman <bill.slo...@ieee.org> wrote:
On Jan 23, 6:20 pm, spflanze <art...@wavenet.org> wrote:
On Jan 21, 11:09 pm, miso <m...@sushi.com> wrote:
If we did a mental exercise and reduced the bits in the DDS, eventually
it would be a square wave. So wouldn't the first noise appear at 3x the
modulation frequency if the modulation frequency and clock rate have an
integer relationship. If so, then your post filter is substantially
easier to build.
That would be true if it were a perfect DAC with infinite bandwidth.
But there is a finite bandwidth and finite slew rate so its output is
other than a succession of perfect squares. Triangular elements are
involved and those have even harmonics.
A triangular wave is just the integral of a square wave, and has the
same - odd only - harmonics, but with the amplitude dropoing in
proportion to the square of the harmonic number.
But yes, the output from a DDS is potentially messier than the output
from simpler switching circuit.
Have you thought about a delay-line filter? You clock your square wave
through a shift register, hook up a suitable resistor to each stage,
and sum the outputs. IIRR a sinc function tapered with a Hamming
window to kills the Gibbs oscillation can give you a very clean sine
wave; you've got residual high-frequency harmonics, but the resistors
can be 0.1% parts (with a bit of padding to get the exact values) and
the lower harmonics can be held to better than 60dB down.
snip
--
Bill Sloman, Nijmegen
It is the sawtooth that has the even harmonics. You are right that the
triangular wave can be considered the result of a square wave with its
higher harmonics being attenuated by an integrating filter.
On Jan 23, 6:20 pm, spflanze <art...@wavenet.org> wrote:
On Jan 21, 11:09 pm, miso <m...@sushi.com> wrote:
If we did a mental exercise and reduced the bits in the DDS, eventually
it would be a square wave. So wouldn't the first noise appear at 3x the
modulation frequency if the modulation frequency and clock rate have an
integer relationship. If so, then your post filter is substantially
easier to build.
That would be true if it were a perfect DAC with infinite bandwidth.
But there is a finite bandwidth and finite slew rate so its output is
other than a succession of perfect squares. Triangular elements are
involved and those have even harmonics.
A triangular wave is just the integral of a square wave, and has the
same - odd only - harmonics, but with the amplitude dropoing in
proportion to the square of the harmonic number.
But yes, the output from a DDS is potentially messier than the output
from simpler switching circuit.
Have you thought about a delay-line filter? You clock your square wave
through a shift register, hook up a suitable resistor to each stage,
and sum the outputs. IIRR a sinc function tapered with a Hamming
window to kills the Gibbs oscillation can give you a very clean sine
wave; you've got residual high-frequency harmonics, but the resistors
can be 0.1% parts (with a bit of padding to get the exact values) and
the lower harmonics can be held to better than 60dB down.
snip
--
Bill Sloman, Nijmegen
It is the sawtooth that has the even harmonics. You are right that the
triangular wave can be considered the result of a square wave with its
higher harmonics being attenuated by an integrating filter.
The sawtooth waveform with its even harmonics is a component of a
staircase waveform:
http://www.analog.com/static/imported-files/tutorials/MT-001.pdf
This is very much like what the sine wave looks like on its way up and
down. The output of the DDS will have even harmonics.
spflanze
Guest
Guest
Wed Jan 25, 2012 6:08 am
On Jan 24, 10:10 am, spflanze <art...@wavenet.org> wrote:
Quote:
On Jan 23, 10:37 pm, spflanze <art...@wavenet.org> wrote:
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
I was wrong about the white noise approximation applying to DDS
quantization noise. Because the quantization noise is synchronous with
the sine wave there is no psuedo randomization and so no white noise.
The harmonics will be clearly defined. My decision not to use the DDS
and instead input a square wave into a switched capacitor filter still
stands because I think I will still get a good result and it will be
easy to generate the square wave from the switching frequency I must
generate anyway. I think the DDS will be an additional part with
little benefit.
John Devereux
Guest
Guest
Wed Jan 25, 2012 9:18 am
spflanze <artist_at_wavenet.org> writes:
Quote:
On Jan 23, 10:37Â pm, spflanze <art...@wavenet.org> wrote:
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
That was me I think.
Overkill for most applications, expensive, power-hungry and awkward to
use. But it is the king of references. It is still unbeaten after nearly
30 years for long term drift and probably noise too.
--
John Devereux
Bill Sloman
Guest
Guest
Wed Jan 25, 2012 9:50 am
On Jan 25, 12:15 am, John Larkin
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
Quote:
On Tue, 24 Jan 2012 15:00:41 -0800 (PST),BillSloman
bill.slo...@ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up.. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
bill.slo...@ieee.org> wrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up.. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
Moron. I never said anything of the kind.
In the thread "OT : so this is what our troops are dying for in
Afghanistan!", on Thu, 17 Aug 2006 08:15:28 -0700, you wrote
"
I have no problem charging as much as I think my customers will be
willing to pay. To keep that number as high as possible, I
differentiate my products by making them insanely good. If they don't
like my stuff, they can buy something else. "
Someone around here s a moron, but it doesn't seem to be me.
Quote:
The real difference between the stuff I design, and the stuff you used
to design, is that most of my stuff sells.
to design, is that most of my stuff sells.
I worked for people who were pushing the envelope at bit further - too
far, as it turned out. If I'd been in a position to do some of the
market research we might have done better, but as it was, it was only
my design skills that were tested, and the stuff that I did get to
build worked.
If this was sci.electronics.market-research you'd have something to
crow about. As it is, you are just posting one more of your irrelevant
insults.
Quote:
And, of course, I'm still designing.
Lucky you.
--
Bill Sloman, Nijmegen
Bill Sloman
Guest
Guest
Wed Jan 25, 2012 10:09 am
On Jan 25, 12:45 am, Phil Hobbs
<pcdhSpamMeSensel...@electrooptical.net> wrote:
Quote:
BillSlomanwrote:
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
You were the one who said that notches were a problem because of their
phase shift,
On Jan 24, 9:05 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 5:22 pm, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 24, 6:29 am, Phil Hobbs
pcdhSpamMeSensel...@electrooptical.net> wrote:
BillSlomanwrote:
On Jan 23, 10:09 pm, "tm" <No_one_h...@white-house.gov> wrote:
"spflanze" <art...@wavenet.org> wrote in message
news:e2aab768-3c58-4381-9701-896e2d3a80fd_at_t13g2000yqg.googlegroups.com...
If both the modulating and demodulating waveforms are square wave
there will be contributions to the output at odd harmonics where the
signal to noise ratios of the system is less. The square wave is the
equivalent of doing both a carrier signal and Fourier Transform at the
fundamental and again at all odd harmonics and summing the results.
The odd harmonics can be filtered. A filter with enough attenuation at
the first of the harmonics and very little attenuation at the carrier
frequency would require many poles, and would require many op amps.
Use a notch (or several as needed) as well as the LPF to get rid of the
harmonics.
Notches aren't a good idea inside a lock-in amplifier set-up. Too much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable.
Actually notches are great inside feedback loops, because their phase
shifts go away at frequencies more than a few notch widths away. Long
long ago, when I was designing atomic force microscopes, we got a factor
of about 4 increase in speed by notching out the mechanical resonance of
the silicon cantilever.
If you have a resonant actuator, a notch filter can be a _big_ win.
Agreed. You just have to want to notch out a frequency that has
nothing to do with your modulation frequency, which doesn't happen all
that often.
Not so--that's exactly what it's good for. You can't notch out the
modulation frequency before the demod, obviously, but you can widen the
lowpass on the output if you notch out the ripple components.
"Widen the lowpass"? You can certainly complicate the post-
demodulation low pass filter by adding a notch at the ripple frequency
- and an easily tunable notch like a bridge differentiator would be
the way to go - but a higher order low pass filter is usually a more
attractive option, since you don't need to tune it to match your
frequency source or worry about temperature drifts moving the notch
away from the ripple frequency.
Nonsense. Higher order filters make the phase shift worse. You're
moving the goal posts.
Actually, you are. Phase shift after demodulation doesn't usually
matter - the exception is when you are closing the loop around the
whole modulator-demodulator set up, and even then you've got to have
some kind of phase-lag budget to make a practicable system.
I know you feel the need to pose as infallible expert, in much the
same way the John Larkin feels the need to insist that all his
products represent "insanely good design", but trying too hard defeats
the purpose.
You were the one who said that notches were a problem because of their
phase shift,
I meant before the demodulator, where the temptation is use a notch
that only lets through the modulation frequency.
You pointed out that you can safely notch out other frequencies -
which is true - but when I agreed you then broadened the discussion to
include the post-demodulator low-pass filter, where a notch could be
useful to take out the ripple at twice the modulation frequency, to
which I responded by pointing that it;s rarely a sensible choice.
Now that we've anatomised the problem in some detail, you now want to
ignore all that detail.
What I actually said was
"Notches aren't a good idea inside a lock-in amplifier set-up. Too
much
phase shift.
You start needing to add all-pass networks to your filter networks to
keep the phase-shift low and stable."
and one could ague that "inside a lock-in amplifier set-up" implies
before the demodulator, though it certainly isn't explicit.
Quote:
and then you try fixing that by bringing in a higher order
filter, which makes the problem worse.
filter, which makes the problem worse.
Not necessarily - whence my mention of all-pass networks.
Quote:
Now all you have left is name calling,
Obviously untrue.
Quote:
and I'll leave that to you.
Hoping that misleading over-simplification and text-chopping are
inoffensive enough not to compromise your "dispassionate expert"
image.
Pity about that.
--
Bill Sloman, Nijmegen
spflanze
Guest
Guest
Thu Jan 26, 2012 7:38 am
On Jan 24, 8:54 pm, John Larkin
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
Quote:
Switched-cap filters are noisy and nasty. They also alias everything
in sight, including their own power supply noise. And they make big
output spikes.
All you need is a modest analog lowpass filter ahead of the ADC, just
to keep the noise down.
John
in sight, including their own power supply noise. And they make big
output spikes.
All you need is a modest analog lowpass filter ahead of the ADC, just
to keep the noise down.
John
I am expecting switching noise. I have seen this in past experience
with them. The frequency of this noise will be an exact integer
multiple of the generated sine wave and so it will be taken out by the
FFT.
The sine wave frequency needs to be adjustable and so then must also
be the filter. Analog filters are not easily adjusted.
The switched capacitor filter I have chosen for the task is the
MAX294. To get the square wave's third harmonic in the notch shown at
1.8 kHz in the frequency response curve on page 4 at:
http://datasheets.maxim-ic.com/en/ds/MAX293-MAX297.pdf
I figure I will need to clock it at 167 times the frequency of the
square wave. The lowest needed frequency is 100 Hz. So the lowest
clock frequency is 16.7kHz. In addition to there being an FFT, I
intend to put a 4.5kHz four pole filter at the output of the MAX294.
This would suppress the 5mVpp clock feed through (page 3) down to
26uV. This and the FFT should be sufficient.
The capacitors in the switched capacitor filter are referenced to
ground. I do not see that there is any direct mechanism to couple
power supply noise to the output. Coupling would be by stray
capacitance and the limitations of the op amp's power supply
rejection. I expect this coupling to be weak. I also intend to use
regulators designed for low noise.
What is left is to find a way to generate all the needed clocks and
square waves. I will start another thread for advice on this.
spflanze
Guest
Guest
Thu Jan 26, 2012 7:43 am
On Jan 25, 12:18 am, John Devereux <j...@devereux.me.uk> wrote:
Quote:
spflanze <art...@wavenet.org> writes:
On Jan 23, 10:37 pm, spflanze <art...@wavenet.org> wrote:
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
That was me I think.
Overkill for most applications, expensive, power-hungry and awkward to
use. But it is the king of references. It is still unbeaten after nearly
30 years for long term drift and probably noise too.
--
John Devereux
On Jan 23, 10:37 pm, spflanze <art...@wavenet.org> wrote:
I am now considering filtering higher harmonics and noise by using a
switched capacitor filter such as described here:http://www.maxim-ic.com/app-notes/index.mvp/id/2081
Instead of inputting a square wave I could input the output of a DDS
and by so doing start with lower harmonics to filter out.
I am concerned about switching noise. In prior experience with
switched capacitor filters I have observed switching noise in the
output. I would like the switching frequency to be well above the
cutoff of the anti-alias filter. For this to be I figure I would need
at least 10,000:1 ratio of switching frequency to cutoff frequency. In
the filters I have seen 100:1 is typical and 1000:1 is rare. I have
not found one with 10,000:1. Does it exist?
Failing that I could make the switching frequency a multiple of 1/T
where T is the length of time the FFT is done over. That would put the
switching frequency in one of the notches in the sync function that is
the frequency response of a Fourier Transform.
I could create a switched capacitor filter out of several chips. But
it would make the board too large. I would need it in a single chip.
There would also be the 6 different switching frequencies to generate.
I know how to do this with several counter and logic chips. But any
lower chip count suggestions are welcome, especially a singe chip with
several channels of output.
I have decided not to use a DDS. I recognize now that the same
quantization noise equation with its white noise approximation that
would apply to the 16 bit ADC would apply to the DDS. So if I use a 10
bit DDS I will get the quantization noise performance of 10 bits, not
16 bits. Since the white noise approximation applies there is no
filtering this out. The 14 Bit DDSs consume too much power. I am going
to input square waves to eight pole elliptical switched capacitor
filter chips. There will be a low level of switching noise that passes
through the LED driver to the photodetector, and through the anti-
aliasing filter. But if the switching frequencies are chosen right the
FFT will notch these frequencies out.
There has been mention in another reply of an LTZ1000 reference. I
have looked at the specs and it looks like a very good one. I plan to
use one them to voltage reference both the ADCs and the square wave
generators. Thanks for bringing this one up.
That was me I think.
Overkill for most applications, expensive, power-hungry and awkward to
use. But it is the king of references. It is still unbeaten after nearly
30 years for long term drift and probably noise too.
--
John Devereux
I changed my mind about it. Its voltage is too high.
elektroda.net NewsGroups Forum Index - Electronics Design - DDS Selection
