Vladimir Vassilevsky wrote:> John Sampson wrote: > > > alexs wrote: > > > >> I need to measure the time it takes signal to turn on and be at steady > >> state frequency (1 to 10MHz)to within 200hz resolution within 2us. The > >> kicker is that I can only sample at 65MSPS. Is this possible? I can > >> get 200hz resolution by sampling for 5ms, but I am only interested in > >> the sample from 0 to 2us, so when I window that section, I loose my > >> resolution. If this problem interests you, I can provide more > >> information on request. > >> > > > > The kicker isn't that you can only sample at 65 Msps, it's that you can > > only use 2 us of data. That constraint limits your resolution to 500 > > kHz, no matter what the sample rate is. > > Very wrong statement.Agreed, for many types of frequency measurement problems.> There is absolutely no relation between the duration of the data and the > frequency resolution. The only thing that matters is the SNR.Isn't the frequency resolution dependent on the product of the duration and the S/N ratio for some signal models? For square waves it seems to be related to the product of the sample rate and the number of square wave cycles in duration. IMHO. YMMV. -- rhn A.T nicholson d.0.t C-o-M
time to frequency measurement
Started by ●November 1, 2006
Reply by ●November 1, 20062006-11-01
Reply by ●November 1, 20062006-11-01
johns@3db-labs.com wrote:> Yes, the instantaneous frequency. How long do I need to look at it for > to know that the signal is within 200 Hz of steady state?Depends on the peak rate of change, the rate of change trajectory, and the noise level (S/N ratio). If the frequency trajectory is consistent enough, you might be able to predict "steady state"+-delta ahead of time from the curvature towards the asymptote. IMHO. YMMV. -- rhn A.T nicholson d.0.t C-o-M
Reply by ●November 1, 20062006-11-01
johns@3db-labs.com writes:> Randy Yates wrote: >> Vladimir Vassilevsky <antispam_bogus@hotmail.com> writes: >> >> > John Sampson wrote: >> > >> >> alexs wrote: >> >> >> >>> I need to measure the time it takes signal to turn on and be at steady >> >>> state frequency (1 to 10MHz)to within 200hz resolution within 2us. The >> >>> kicker is that I can only sample at 65MSPS. Is this possible? I can >> >>> get 200hz resolution by sampling for 5ms, but I am only interested in >> >>> the sample from 0 to 2us, so when I window that section, I loose my >> >>> resolution. If this problem interests you, I can provide more >> >>> information on request. >> >>> >> >> The kicker isn't that you can only sample at 65 Msps, it's that you >> >> can only use 2 us of data. That constraint limits your resolution to >> >> 500 kHz, no matter what the sample rate is. >> > >> > Very wrong statement. >> > >> > There is absolutely no relation between the duration of the data and >> > the frequency resolution. The only thing that matters is the SNR. >> >> I agree with Vladimir. >> >> Consider this gedanken. You have a pure sine wave that's been sampled and >> input to a Hilbert transformer since the beginning of time. You select >> two [complex] samples, 0*T and 1*T, from this infinite stream and compute >> >> f = (arctan2(x1.y/x1.x) - arctan2(x0.y/x0.x)) / T. >> >> This frequency is perfect with just two samples if there is no noise. >> -- >> % Randy Yates % "Though you ride on the wheels of tomorrow, >> %% Fuquay-Varina, NC % you still wander the fields of your >> %%% 919-577-9882 % sorrow." >> %%%% <yates@ieee.org> % '21st Century Man', *Time*, ELO >> http://home.earthlink.net/~yatescr > > Yes, the instantaneous frequency. How long do I need to look at it for > to know that the signal is within 200 Hz of steady state?I'm presuming there is a signal that is monotonically increasing or decreasing in frequency, asymptotically approaching some steady-state frequency. If you know a-priori how the signal approaches the asymptote, and can model it mathematically, then you can take a few instantaneous frequency measurements and do a least-squares fit to your parametric model. The resulting model parameters will tell you when you're within 200 Hz of the steady-state. I think. This is off the top of my head, and it's getting pretty sparse up there... -- % Randy Yates % "Ticket to the moon, flight leaves here today %% Fuquay-Varina, NC % from Satellite 2" %%% 919-577-9882 % 'Ticket To The Moon' %%%% <yates@ieee.org> % *Time*, Electric Light Orchestra http://home.earthlink.net/~yatescr
Reply by ●November 1, 20062006-11-01
johns@3db-labs.com wrote:> Randy Yates wrote: >> Vladimir Vassilevsky <antispam_bogus@hotmail.com> writes: >> >>> John Sampson wrote: >>> >>>> alexs wrote: >>>> >>>>> I need to measure the time it takes signal to turn on and be at steady >>>>> state frequency (1 to 10MHz)to within 200hz resolution within 2us. The >>>>> kicker is that I can only sample at 65MSPS. Is this possible? I can >>>>> get 200hz resolution by sampling for 5ms, but I am only interested in >>>>> the sample from 0 to 2us, so when I window that section, I loose my >>>>> resolution. If this problem interests you, I can provide more >>>>> information on request. >>>>> >>>> The kicker isn't that you can only sample at 65 Msps, it's that you >>>> can only use 2 us of data. That constraint limits your resolution to >>>> 500 kHz, no matter what the sample rate is. >>> Very wrong statement. >>> >>> There is absolutely no relation between the duration of the data and >>> the frequency resolution. The only thing that matters is the SNR. >> I agree with Vladimir. >> >> Consider this gedanken. You have a pure sine wave that's been sampled and >> input to a Hilbert transformer since the beginning of time. You select >> two [complex] samples, 0*T and 1*T, from this infinite stream and compute >> >> f = (arctan2(x1.y/x1.x) - arctan2(x0.y/x0.x)) / T. >> >> This frequency is perfect with just two samples if there is no noise. >> -- >> % Randy Yates % "Though you ride on the wheels of tomorrow, >> %% Fuquay-Varina, NC % you still wander the fields of your >> %%% 919-577-9882 % sorrow." >> %%%% <yates@ieee.org> % '21st Century Man', *Time*, ELO >> http://home.earthlink.net/~yatescr > > Yes, the instantaneous frequency. How long do I need to look at it for > to know that the signal is within 200 Hz of steady state?At least until you know what steady state will turn out to be. Jerry -- "The rights of the best of men are secured only as the rights of the vilest and most abhorrent are protected." - Chief Justice Charles Evans Hughes, 1927 ���������������������������������������������������������������������
Reply by ●November 1, 20062006-11-01
"Ron N." <rhnlogic@yahoo.com> writes:> IMHO. YMMV.Ron, what does "YMMV" stand for? -- % Randy Yates % "With time with what you've learned, %% Fuquay-Varina, NC % they'll kiss the ground you walk %%% 919-577-9882 % upon." %%%% <yates@ieee.org> % '21st Century Man', *Time*, ELO http://home.earthlink.net/~yatescr
Reply by ●November 1, 20062006-11-01
Randy Yates wrote:> johns@3db-labs.com writes: > > > Randy Yates wrote: > >> Vladimir Vassilevsky <antispam_bogus@hotmail.com> writes: > >> > >> > John Sampson wrote: > >> > > >> >> alexs wrote: > >> >> > >> >>> I need to measure the time it takes signal to turn on and be at steady > >> >>> state frequency (1 to 10MHz)to within 200hz resolution within 2us. The > >> >>> kicker is that I can only sample at 65MSPS. Is this possible? I can > >> >>> get 200hz resolution by sampling for 5ms, but I am only interested in > >> >>> the sample from 0 to 2us, so when I window that section, I loose my > >> >>> resolution. If this problem interests you, I can provide more > >> >>> information on request. > >> >>> > >> >> The kicker isn't that you can only sample at 65 Msps, it's that you > >> >> can only use 2 us of data. That constraint limits your resolution to > >> >> 500 kHz, no matter what the sample rate is. > >> > > >> > Very wrong statement. > >> > > >> > There is absolutely no relation between the duration of the data and > >> > the frequency resolution. The only thing that matters is the SNR. > >> > >> I agree with Vladimir. > >> > >> Consider this gedanken. You have a pure sine wave that's been sampled and > >> input to a Hilbert transformer since the beginning of time. You select > >> two [complex] samples, 0*T and 1*T, from this infinite stream and compute > >> > >> f = (arctan2(x1.y/x1.x) - arctan2(x0.y/x0.x)) / T. > >> > >> This frequency is perfect with just two samples if there is no noise. > >> -- > >> % Randy Yates % "Though you ride on the wheels of tomorrow, > >> %% Fuquay-Varina, NC % you still wander the fields of your > >> %%% 919-577-9882 % sorrow." > >> %%%% <yates@ieee.org> % '21st Century Man', *Time*, ELO > >> http://home.earthlink.net/~yatescr > > > > Yes, the instantaneous frequency. How long do I need to look at it for > > to know that the signal is within 200 Hz of steady state? > > I'm presuming there is a signal that is monotonically increasing or > decreasing in frequency, asymptotically approaching some steady-state > frequency. > > If you know a-priori how the signal approaches the asymptote,In fact, it seems like you would have to know how it approaches the asymptote. If the frequency can overshoot once or more, then even a measurement showing that the frequency is dead-on in some window might not be a good indication that the frequency has reached steady-state. It might just be passing through. IMHO. YMMV. -- rhn A.T nicholson d.0.t C-o-M
Reply by ●November 1, 20062006-11-01
Vladimir Vassilevsky wrote:> johns@3db-labs.com wrote: > > > > > > http://www.bores.com/courses/intro/basics/1_resol.htm > > > > Bullshit. > > VLVHa-Ha-Ha-Ha-Ha-....... Second that !
Reply by ●November 1, 20062006-11-01
On Thu, 02 Nov 2006 03:18:38 +0000, Randy Yates wrote:> "Ron N." <rhnlogic@yahoo.com> writes: > >> IMHO. YMMV. > > Ron, what does "YMMV" stand for?Generally thought to be "Your Milage May Vary". Sort of internet-ese for "Caveat Emptor". Cheers, -- Andrew
Reply by ●November 2, 20062006-11-02
"Ron N." <rhnlogic@yahoo.com> writes:> Randy Yates wrote: >> johns@3db-labs.com writes: >> >> > Randy Yates wrote: >> >> Vladimir Vassilevsky <antispam_bogus@hotmail.com> writes: >> >> >> >> > John Sampson wrote: >> >> > >> >> >> alexs wrote: >> >> >> >> >> >>> I need to measure the time it takes signal to turn on and be at steady >> >> >>> state frequency (1 to 10MHz)to within 200hz resolution within 2us. The >> >> >>> kicker is that I can only sample at 65MSPS. Is this possible? I can >> >> >>> get 200hz resolution by sampling for 5ms, but I am only interested in >> >> >>> the sample from 0 to 2us, so when I window that section, I loose my >> >> >>> resolution. If this problem interests you, I can provide more >> >> >>> information on request. >> >> >>> >> >> >> The kicker isn't that you can only sample at 65 Msps, it's that you >> >> >> can only use 2 us of data. That constraint limits your resolution to >> >> >> 500 kHz, no matter what the sample rate is. >> >> > >> >> > Very wrong statement. >> >> > >> >> > There is absolutely no relation between the duration of the data and >> >> > the frequency resolution. The only thing that matters is the SNR. >> >> >> >> I agree with Vladimir. >> >> >> >> Consider this gedanken. You have a pure sine wave that's been sampled and >> >> input to a Hilbert transformer since the beginning of time. You select >> >> two [complex] samples, 0*T and 1*T, from this infinite stream and compute >> >> >> >> f = (arctan2(x1.y/x1.x) - arctan2(x0.y/x0.x)) / T. >> >> >> >> This frequency is perfect with just two samples if there is no noise. >> >> -- >> >> % Randy Yates % "Though you ride on the wheels of tomorrow, >> >> %% Fuquay-Varina, NC % you still wander the fields of your >> >> %%% 919-577-9882 % sorrow." >> >> %%%% <yates@ieee.org> % '21st Century Man', *Time*, ELO >> >> http://home.earthlink.net/~yatescr >> > >> > Yes, the instantaneous frequency. How long do I need to look at it for >> > to know that the signal is within 200 Hz of steady state? >> >> I'm presuming there is a signal that is monotonically increasing or >> decreasing in frequency, asymptotically approaching some steady-state >> frequency. >> >> If you know a-priori how the signal approaches the asymptote, > > In fact, it seems like you would have to know how it approaches > the asymptote. If the frequency can overshoot once or more, > then even a measurement showing that the frequency is dead-on > in some window might not be a good indication that the frequency > has reached steady-state. It might just be passing through.My "monotonic" presumption precludes that scenario, but your point is still valid, Ron. Without knowing the behavior, the trajectory could potentially do all kinds of strange things. I wonder what the OP's requirement really is. -- % Randy Yates % "...the answer lies within your soul %% Fuquay-Varina, NC % 'cause no one knows which side %%% 919-577-9882 % the coin will fall." %%%% <yates@ieee.org> % 'Big Wheels', *Out of the Blue*, ELO http://home.earthlink.net/~yatescr
Reply by ●November 2, 20062006-11-02
Vladimir Vassilevsky wrote:> > > johns@3db-labs.com wrote: > > >> >> http://www.bores.com/courses/intro/basics/1_resol.htm >> > > Bullshit. > > VLV > >The inverse relationship between frequency resolution and observation time normally relates to nonparametric signal models. The OP didn't say much about the signal so I assumed the problem was to tell when an arbitrary signal stops moving around in frequency to within 200 Hz. For example it could be a burst modem where the modulation turns on too soon after the LO does. In order to tell if the PSD of the signal is holding still to within 200 Hz it seems to me that you need at least 200 Hz frequency resolution. If we know something about the signal, which the OP most likely does, then we can draw from a large bag of tricks. John
