Ultra low frequency VCO
in [Design]
|
Prev: AofE 3rd Edition
Next: Labelling Prototypes
From: Phil Hobbs on 14 Jul 2010 10:11 On 7/14/2010 12:49 AM, JosephKK wrote: > On Mon, 12 Jul 2010 23:23:56 -0400, Phil Hobbs > <pcdhSpamMeSenseless(a)electrooptical.net> wrote: > >> JosephKK wrote: >>> On Mon, 12 Jul 2010 09:37:07 -0400, Phil Hobbs >>> <pcdhSpamMeSenseless(a)electrooptical.net> wrote: >>> >>>> JosephKK wrote: >>>>> On Fri, 9 Jul 2010 10:22:34 -0700 (PDT), j<jdc1789(a)gmail.com> wrote: >>>>> >>>>>> Resolution of noise vs frequency, (as in bw), is the issue in phase >>>>>> noise measurements. The OP never stated the offset from the carrier >>>>>> nor bandwidth. Or maybe I just missed it. >>>>>> >>>>>> It's not clear to me why JosephKK thinks this would be either a time >>>>>> consuming or difficult measurement to make. Assuming the appropriate >>>>>> measurement system is in hand 100 dBc numbers are easily achievable. >>>>>> Whether it's 60 Hz or several GHz's the global issues are the same in >>>>>> making a phase noise measurement. >>>>>> >>>>>> But having said the above, without the OP responding I guess it really >>>>>> doesn't matter. But I'd like to know more about the application and >>>>>> derive solutions from there. >>>>> >>>>> OK. For a carrier of 60 MHz. Pick an instrument or test setup of your >>>>> choice, state the model[s]. Clearly explain just what is going on in the >>>>> measurement and the time it takes to accumulate sufficient data to make >>>>> the measurement. Explain why it takes that much time to reach a reliable >>>>> measurement of -100 dBc phase noise at that carrier frequency. >>>>> >>>>> Now see how well it scales to one million times lower fundamental >>>>> frequency without a similar scaling in measurement time. >>>> It's the modulation frequency that's relevant, not the carrier >>>> frequency. Measurements get slower when you reduce the bandwidth. >>>> >>>> (You can see why this doesn't work if you imagine running it >>>> backwards--mixing or multiplying up to some very high frequency doesn't >>>> allow you to make a measurement with 1 Hz bandwidth any faster. >>>> >>>> >>>> Cheers >>>> >>>> Phil Hobbs >>> >>> Now what is the equivalent bandwidth of -100 dBc for a 60 Hz carrier? >>> Since you said 20 log() basis 60 * 10^-5 is 600 microHz. That would have >>> to take some minutes, and if you wanted a proper 10 to 1 measurement >>> buffer, it takes ten times longer. Call it 10,000 seconds? A few hours. >>> And the reference stability etc., i remarked on is coming into play. >> >> >> You're confused, I'm afraid. -100 dBc phase noise in a given bandwidth >> (say 1 Hz, but it doesn't matter) is 7 microradians RMS. Using a 5V >> swing and a CMOS analogue gate as a phase detector, that's >> >> dV = 7e-6 rad RMS * 5V/(pi rad) = 11 microvolts RMS, >> >> which is trivial to measure in a 1 Hz bandwidth in a few seconds--it's >> 80 dB above the noise of a good op amp, so you just have to wait for the >> filter to settle. >> >> Cheers >> >> Phil Hobbs > > OK. I have your "Making it All Work" and AoE 2nd Ed and more. Where do > i go to get less confused? This phase noise measurement is twisted. If you read the derivation in Section 13.6 and do the math, which isn't difficult--just sums and differences of trig functions--we should be talking the same language. The main point is that we discuss small-amplitude phase noise using the small angle approximation, i.e. sin theta ~= theta, so that it's just like amplitude noise except that it's in phase quadrature with the carrier. That makes it a bog-standard propagation-of-errors calculation: you take all the noise sources, multiply them by the relevant partial derivatives, and compute the RMS sum. If you add white noise, half winds up in the I phase, which looks like amplitude variations, and half in the Q phase, which looks like phase variations. The small angle behaviour makes the statistics and frequency spectrum of the resulting phase and amplitude noise equal to those of the original additive noise. It's quite pretty. When the SNR is below about 20 dB, we have to start being a lot more careful mathematically. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal ElectroOptical Innovations 55 Orchard Rd Briarcliff Manor NY 10510 845-480-2058 hobbs at electrooptical dot net http://electrooptical.net
From: Phil Hobbs on 14 Jul 2010 11:08 On 7/13/2010 11:58 PM, j wrote: > On Jul 13, 7:41 pm, Phil Hobbs > <pcdhSpamMeSensel...(a)electrooptical.net> wrote: >> j wrote: >>>> Genuine phase noise sidebands have flat tops, so they aren't as >>>> sensitive to modulation frequency as FM noise. Various authors go to >>>> various lengths in trying to identify regions where the noise goes as >>>> 1/f, 1/f**2,.... >> >>> What in the world are you saying? Sounds kind of ignorant to me � but >>> I�ll reserve judgment until you answer. >> >>> I made a living at designing multiloop uw / rf synthesizers and taking >>> this statement as fact sure wouldn�t have helped. >> >>> regards >> >> Which statement? You don't think that pure phase noise is white? Or >> that different authors say different things? >> >> I plead guilty to ignorance of many things. >> > > Yeah, I suspected you were talking about white noise. > > Unfortunately for folks that design low noise freq synthesizers white > noise isn�t the tough spot. We typically live in those 1/f places. > The whole process is about shaping that noise profile. Without > targeted system spec�s, one can see why it�s virtually impossible to > select loop components such as a VCO�s, amps, etc., for this type of > job. > Understood. My first engineering job was designing 2/3 of the time and frequency reference boards for the first direct-broadcast satellite system (the Spacetel system from AEL Microtel), including the VHF synthesizer that controlled the 14 GHz local oscillator on both the central station and the remotes. The noise spec was 7 Hz RMS in 5-100 Hz bandwidth around a 14 GHz carrier, i.e. after being multiplied up by 120 times from the output of my board. I was allocated half of this budget, i.e. 5 Hz RMS at 14 GHz, or 0.041 Hz at 115 MHz, and my synthesizer had to be tunable over about 5 MHz in steps of 8.3333... kHz (1 MHz at the LO frequency). This was in 1981-83, remember, which was well before DDS. I had no idea how hard that was before I started--I had a brand new astronomy and physics B.Sc., and only a hobby background in electronics (though I did start when I was 10 years old). I knew about PLLs, but I'd never seen one, let alone designed one. Talk about being chucked in the deep end of the pool! (I eventually made a fairly novel fractional-N synthesizer with an 833.33... kHz comparison frequency, using a MC12013 ECL 10/11 dual modulus prescaler(*) with the modulus controlled by a 74LS163 synchronous decade counter, whose carry input was driven by a string of CD4527 CMOS BCD rate multipliers. The rate multiplier jitter was pretty well outside the LO PLL's bandwidth, and certainly wasn't in the 5-100 Hz band that they mostly cared about. One nice feature was that you didn't need a switch setting table--because it was BCD and used a synchronous divide-by-12 to get the reference, you just set the BCD DIP switches to the desired 14 GHz LO frequency. It all eventually worked fine, thanks to Floyd Gardner's book and the Mini-Circuits catalogue. Its main wart was that even with lead-lag compensation, I _still_ couldn't get enough loop gain to control the noise of an LC VCO with a 5 MHz tuning range. I eventually had to retreat and use a VCXO, meaning that you had to change a crystal as well as set the DIP switches. (Changing crystals was SOP in those days, so nobody minded too much.) The oscillator was a one-transistor Colpitts with self-limiting, just like in the ARRL Handbook. If I'd known how to design better oscillators, or had had coaxial resonators, I could probably have kept real tunability. Such is life. > Btw, I apologize for the ignorant comment � didn�t mean to sound so > nasty. I regret that. No worries. The plus side of having large areas of ignorance is the opportunity to learn lots of new things, which is one reason I like SED in spite of the spam, troll baiting, and flame wars. I think that standing at a white board arguing about technical stuff with a few smart people is the most fun you can have standing up. (I have a directory full of pithy Usenet posts that I refer to periodically--some of the stuff in my book came out of things I learned here.) Cheers Phil Hobbs (*) Did you design that one, Jim? -- Dr Philip C D Hobbs Principal ElectroOptical Innovations 55 Orchard Rd Briarcliff Manor NY 10510 845-480-2058 hobbs at electrooptical dot net http://electrooptical.net
From: George Herold on 14 Jul 2010 12:46 On Jul 14, 12:49 am, "JosephKK"<quiettechb...(a)yahoo.com> wrote: > On Mon, 12 Jul 2010 23:23:56 -0400, Phil Hobbs > > > > > > <pcdhSpamMeSensel...(a)electrooptical.net> wrote: > >JosephKK wrote: > >> On Mon, 12 Jul 2010 09:37:07 -0400, Phil Hobbs > >> <pcdhSpamMeSensel...(a)electrooptical.net> wrote: > > >>> JosephKK wrote: > >>>> On Fri, 9 Jul 2010 10:22:34 -0700 (PDT), j <jdc1...(a)gmail.com> wrote: > > >>>>> Resolution of noise vs frequency, (as in bw), is the issue in phase > >>>>> noise measurements. The OP never stated the offset from the carrier > >>>>> nor bandwidth. Or maybe I just missed it. > > >>>>> Its not clear to me why JosephKK thinks this would be either a time > >>>>> consuming or difficult measurement to make. Assuming the appropriate > >>>>> measurement system is in hand 100 dBc numbers are easily achievable.. > >>>>> Whether its 60 Hz or several GHzs the global issues are the same in > >>>>> making a phase noise measurement. > > >>>>> But having said the above, without the OP responding I guess it really > >>>>> doesnt matter. But Id like to know more about the application and > >>>>> derive solutions from there. > > >>>> OK. For a carrier of 60 MHz. Pick an instrument or test setup of your > >>>> choice, state the model[s]. Clearly explain just what is going on in the > >>>> measurement and the time it takes to accumulate sufficient data to make > >>>> the measurement. Explain why it takes that much time to reach a reliable > >>>> measurement of -100 dBc phase noise at that carrier frequency. > > >>>> Now see how well it scales to one million times lower fundamental > >>>> frequency without a similar scaling in measurement time. > >>> It's the modulation frequency that's relevant, not the carrier > >>> frequency. Measurements get slower when you reduce the bandwidth. > > >>> (You can see why this doesn't work if you imagine running it > >>> backwards--mixing or multiplying up to some very high frequency doesn't > >>> allow you to make a measurement with 1 Hz bandwidth any faster. > > >>> Cheers > > >>> Phil Hobbs > > >> Now what is the equivalent bandwidth of -100 dBc for a 60 Hz carrier? > >> Since you said 20 log() basis 60 * 10^-5 is 600 microHz. That would have > >> to take some minutes, and if you wanted a proper 10 to 1 measurement > >> buffer, it takes ten times longer. Call it 10,000 seconds? A few hours. > >> And the reference stability etc., i remarked on is coming into play. > > >You're confused, I'm afraid. -100 dBc phase noise in a given bandwidth > >(say 1 Hz, but it doesn't matter) is 7 microradians RMS. Using a 5V > >swing and a CMOS analogue gate as a phase detector, that's > > >dV = 7e-6 rad RMS * 5V/(pi rad) = 11 microvolts RMS, > > >which is trivial to measure in a 1 Hz bandwidth in a few seconds--it's > >80 dB above the noise of a good op amp, so you just have to wait for the > >filter to settle. > > >Cheers > > >Phil Hobbs > > OK. I have your "Making it All Work" and AoE 2nd Ed and more. Where do > i go to get less confused? This phase noise measurement is twisted.- Hide quoted text - > > - Show quoted text - Hi Joseph, I'm trying to get my head around this too. (I like Phil's intro to section 13.6, "We live in a fallen world, so the signals we process are never free of noise, distortion, and extraneous interfering signals.")* I think it would help me if I understood how one measures the phase noise. My simple minded approach would be to trigger my digital 'scope on the carrier zero crossing, and then look 'down stream' 100 or 1,000 periods later and see how much 'jitter' there was in the delayed zero crossing. Seems like there must be a better way. George H. *Does this mean there is no noise in heaven? (all R's have zero temperature)
From: Joel Koltner on 14 Jul 2010 12:49 Phil Hobbs wrote: > Understood. My first engineering job was designing 2/3 of the time > and frequency reference boards for the first direct-broadcast > satellite system (the Spacetel system from AEL Microtel), including > the VHF synthesizer that controlled the 14 GHz local oscillator on > both the central station and the remotes. If you don't mind my asking, Phil, how long did that all take? With the synthesizers-in-a-chip (PLL+VCO), off-the-shelf VCOs, plus the design tools from Analog Devices, etc., I have a suspicion that many such RF generators are now given all of perhaps a day or two of design time. :-) ---Joel
From: Phil Hobbs on 14 Jul 2010 14:25
Joel Koltner wrote: > Phil Hobbs wrote: >> Understood. My first engineering job was designing 2/3 of the time >> and frequency reference boards for the first direct-broadcast >> satellite system (the Spacetel system from AEL Microtel), including >> the VHF synthesizer that controlled the 14 GHz local oscillator on >> both the central station and the remotes. > > If you don't mind my asking, Phil, how long did that all take? > > With the synthesizers-in-a-chip (PLL+VCO), off-the-shelf VCOs, plus the > design tools from Analog Devices, etc., I have a suspicion that many > such RF generators are now given all of perhaps a day or two of design > time. :-) > > ---Joel > I joined Microtel in about June of 1981, and left in August 1983 to get married and go to grad school. The first few months were working on the system demo, and the rest of it was spent getting the Pilot Tone Generator and Timing & Frequency Unit designed, breadboarded (no SPICE either), laid out, and tested with the 14 GHz LO setup. Plus miscellaneous other stuff. So I'd say a year, give or take. I had a fair few false starts in there, too, of course--such as trying to use anything with an 8 kHz comparison frequency! Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal ElectroOptical Innovations 55 Orchard Rd Briarcliff Manor NY 10510 845-480-2058 hobbs at electrooptical dot net http://electrooptical.net |