Hi If the noise is "known flat" it's a good way to check system bandwidth. Some means of checking response is indeed very necessary. Not switching the preamp is indeed a good thing. The sound card does not have the range of the 3561, so with the sound card the beat note absolutely require a switch. The switch adds a second calibration step at audio. To calibrate level, the levels of both the noise and carrier need to be well known. For a "basement system" the measure the slope as it crosses zero is likely more accurate. That of course assumes that you do the audio gain and response stuff properly. Many of the free applications that are out there will put a tone out of the card and track it back into the card. That should at least provide a tone to work with. I also should be something that can be fairly easily verified. The issue of mixer output impedance is still a little tricky without RF noise loading. --------------------------- An audio noise source with good low frequency performance is a good thing for checking the lock box induced stuff. If I can do that with the sound card - great. If I have to do it with something like a PWM out of a PIC - not a lot of money there. Not even a lot of money with a PIC driving a 16 bit DAC. Bob On Feb 18, 2010, at 8:52 PM, Bruce Griffiths wrote: > A better calibration technique may be to add a known amount of RF noise (eg from an amplified noise diode or similar) to one of the mixer inputs. > This allows calibration without switching the sound card preamp gain to a value different from that used to make measurements. > The added RF noise only has to be white over the spectrum analyser bandwidth. > This allows the frequency response of the entire measurement setup to be determined. > > Bruce > > Bob Camp wrote: >> Hi >> >> I did a search for *input* DC coupled sound cards and came up empty handed. There are *output* DC coupled cards out there. >> >> I have a pile of old Audigy 2's that are ripe for destruction. Essentially they are free at this point in their life. I've also got a E-MU 0202, an Xonar DX, a couple of Presonus firewire boxes running around. As the price goes up, the probability of destructive mods goes down pretty fast. The Xonar is probably the one that will be in the "final product". >> >> ----------- >> >> The lock box provides an easy way to bump one source off far enough for a beat note calibration. Having a DC coupled "scope" mode would be very nice for doing an accurate calibration. Sine wave approximations are one step to many. >> >> --------- >> >> You could model the input to the sound card. I'll certainly try that on some of the more expensive boxes. Also going to 1 Hz is not an every day need. Most of these cards should do just fine at 10 Hz with no mods. >> >> -------- >> >> Still looking for software. >> >> Bob >> >> >> On Feb 18, 2010, at 8:31 PM, John Miles wrote: >> >> >>> Got it. So you're planning to have a way to supply an offset frequency to >>> generate a beatnote for calibration at some known reduced amplitude, >>> correct? That, and an FFT program that understands how to normalize to 1 Hz >>> BW (and correct the window's noise-equivalent bandwidth) should be all you >>> need. Bonus if the FFT program allows the calibration factor to be added to >>> whatever numbers it draws on the Y-axis. >>> >>> What brand/model sound card are you looking to use? With a white-noise >>> source you might be able to add a calibration step that takes the effect of >>> the input coupling capacitor out of the measurement, so you wouldn't need to >>> modify the card itself. There was an article in QEX a year or so ago on the >>> basic idea. Again, this all assumes you have some control over the >>> software. >>> >>> There may also be some DC-coupled sound cards at that end of the market, >>> too -- I haven't looked lately. >>> >>> -- john, KE5FX >>> >>> >>>> -----Original Message----- >>>> From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com]On >>>> Behalf Of Bob Camp >>>> Sent: Thursday, February 18, 2010 5:10 PM >>>> To: Discussion of precise time and frequency measurement >>>> Subject: Re: [time-nuts] Sound Card Spectrum Analyzer >>>> >>>> >>>> Hi >>>> >>>> Ok, A bit more info: >>>> >>>> 1) Quadrature PLL using an RPD-1 DBM and a home brew lock box. >>>> 2) Willingness to accept that I'm measuring a pair of oscillators >>>> 3) Plenty of sources at the appropriate frequencies >>>> 4) First took a shot at this in 1975 (I forget the Fluke app note >>>> number ...) been doing it ever since >>>> 5) Appropriate preamp between the RPD and sound card is a work in progress >>>> 6) Sound card is a 192 KHz / 24 bit / ~110 db snr class card >>>> 7) Sound card will get butchered for the application. >>>> 8) Should be able to hit -165 ish floor, -120 ish at 1 Hz >>>> >>>> Except for the 16 bit limitation, Baudline looks like it's got >>>> what I need. >>>> >>>> Bob >>>> >>>> >>>> On Feb 18, 2010, at 7:53 PM, John Miles wrote: >>>> >>>> >>>>> Unfortunately there's no way to build a sound-card application that can >>>>> measure phase noise in the general case without a lot of additional >>>>> hardware. Baseband PN analysis with an FFT presupposes that >>>>> >>>> you have some >>>> >>>>> external means of downconverting the DUT signal to DC with a superior >>>>> reference at the same frequency, tuned with a quadrature PLL. >>>>> >>>> There must be >>>> >>>>> a suitable high-pass filter and LNA to block any DC residuals >>>>> >>>> and preamplify >>>> >>>>> the remaining part of the noise sideband. >>>>> >>>>> Further, it's often the case that noise close to the carrier is strong >>>>> enough to keep you from being able to use enough gain to see >>>>> >>>> the broadband >>>> >>>>> floor, so you actually need more than one high-pass filter ahead of the >>>>> sound card in many cases. These switchable filters were >>>>> >>>> mandatory with the >>>> >>>>> old 13-bit signal analyzers like the 3561A, and may still be >>>>> >>>> needed today if >>>> >>>>> you want to look down to 1 Hz. If you restrict your offset >>>>> >>>> range to (say) >>>> >>>>> 100 Hz to 20 kHz and require a 24-bit sound card, you can >>>>> >>>> probably get away >>>> >>>>> without the switchable HPFs. >>>>> >>>>> It'd be helpful to know exactly what sorts of measurements you >>>>> >>>> need to make, >>>> >>>>> and on what devices. PN measurement is a *lot* of work, on both the >>>>> software and hardware sides. Much of it goes into developing a suitable >>>>> calibration process. Take a look at the 3048A manuals >>>>> >>>> sometime, realizing >>>> >>>>> that the 3048A hardware itself is not very complicated... >>>>> >>>>> -- john, KE5FX >>>>> >>>>> >>>>>> -----Original Message----- >>>>>> From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com]On >>>>>> Behalf Of Bob Camp >>>>>> Sent: Thursday, February 18, 2010 4:18 PM >>>>>> To: Discussion of precise time and frequency measurement >>>>>> Subject: Re: [time-nuts] Sound Card Spectrum Analyzer >>>>>> >>>>>> >>>>>> Hi >>>>>> >>>>>> Both are very cool programs. Both are full of all sorts of neat >>>>>> features. As far as I can see, neither one has the features I'm after. >>>>>> >>>>>> More or less - I want it to run like a clunky HP audio analyzer >>>>>> rather than a very cool tool for ham radio. >>>>>> >>>>>> Bob >>>>>> >>>>>> On Feb 18, 2010, at 6:33 PM, Don Latham wrote: >>>>>> >>>>>> >>>>>>> You have looked at: >>>>>>> spectran and spectrum lab ? >>>>>>> Don >>>>>>> >>>>>>> Bob Camp >>>>>>> >>>>>>>> Hi >>>>>>>> >>>>>>>> Assuming I have a decent sound card, and a computer, the next >>>>>>>> >>>>>> thing I need >>>>>> >>>>>>>> is software. If I want: >>>>>>>> >>>>>>>> Required: >>>>>>>> >>>>>>>> 1) non- commercial >>>>>>>> 2) 1 Hz normalization >>>>>>>> 3) good low frequency processing (decimation ahead of the fft) >>>>>>>> 4) low cost >>>>>>>> >>>>>>>> Much preferred: >>>>>>>> >>>>>>>> 5) a non-evil OS >>>>>>>> 6) Rational performance on a non-quad core system >>>>>>>> 7) free >>>>>>>> 8) rational calibration >>>>>>>> 9) scope view. >>>>>>>> 10) reasonable graphics >>>>>>>> 11) active support by the author >>>>>>>> >>>>>>>> The application is measuring phase noise. That what makes 2& >>>>>>>> >>>>>> 3 pop up on >>>>>> >>>>>>>> the list. >>>>>>>> >>>>>>>> I've looked at a lot of programs and they all seem to be >>>>>>>> >>>>>> pretty slick. The >>>>>> >>>>>>>> ones I've looked at so far don't quite hit the mark for phase >>>>>>>> >>>>>> noise. I'm >>>>>> >>>>>>>> pretty sure that there are others on the list who have dug >>>>>>>> >>>>>> into this same >>>>>> >>>>>>>> issue already. >>>>>>>> >>>>>>>> Bob >>>>>>>> _______________________________________________ >>>>>>>> time-nuts mailing list -- time-nuts@febo.com >>>>>>>> To unsubscribe, go to >>>>>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>>>>>>> and follow the instructions there. >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Dr. Don Latham AJ7LL >>>>>>> Six Mile Systems LLP >>>>>>> 17850 Six Mile Road >>>>>>> POB 134 >>>>>>> Huson, MT, 59846 >>>>>>> VOX 406-626-4304 >>>>>>> www.lightningforensics.com >>>>>>> www.sixmilesystems.com >>>>>>> >>>>>>> >>>>>>> _______________________________________________ >>>>>>> time-nuts mailing list -- time-nuts@febo.com >>>>>>> To unsubscribe, go to >>>>>>> >>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>>>> >>>>>> and follow the instructions there. >>>>>> >>>>>> >>>>> >>>>> _______________________________________________ >>>>> time-nuts mailing list -- time-nuts@febo.com >>>>> To unsubscribe, go to >>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>>>> and follow the instructions there. >>>>> >>>>> >>>>> _______________________________________________ >>>>> time-nuts mailing list -- time-nuts@febo.com >>>>> To unsubscribe, go to >>>>> >>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>> >>>> and follow the instructions there. >>>> >>>> >>> >>> _______________________________________________ >>> time-nuts mailing list -- time-nuts@febo.com >>> To unsubscribe, go to >>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>> and follow the instructions there. >>> >>> >>> _______________________________________________ >>> time-nuts mailing list -- time-nuts@febo.com >>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>> and follow the instructions there. >>> >>> >> >> _______________________________________________ >> time-nuts mailing list -- time-nuts@febo.com >> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >> and follow the instructions there. >> >> > > > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. >

Hi All that and open source code as well. Here comes the sidereal nixie wrist watch .... Bob On Feb 18, 2010, at 9:06 PM, David Forbes wrote: > Fellow time-nuts, > > Please forgive the tangential and blatantly commercial subject matter. I remember that a few of you expressed interest in my nixie tube wristwatch when I last mentioned it. > > I have finally gotten it back into production after a too-long hiatus brought on by obsolete parts. > > http://www.nixiewatch.com/ > > And no, it doesn't have a rubidium oscillator. But you can adjust the crystal oscillator frequency (test point provided) and obsess over its inadequacies as a precision time source. And the Woz wears one. > -- > > --David Forbes, Tucson, AZ > http://www.cathodecorner.com/ > > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. >

> > If the noise is "known flat" it's a good way to check system > bandwidth. Some means of checking response is indeed very necessary. > > Not switching the preamp is indeed a good thing. The sound card > does not have the range of the 3561, so with the sound card the > beat note absolutely require a switch. The switch adds a second > calibration step at audio. Maybe not -- the 3561A has nowhere near the dynamic range of the sound card (13 bits versus 20+, or about 80 dB versus ~117 dB for a good sound card). Normally you do beatnote calibration by injecting a signal at (level of DUT - gain of LNA), or 40 dB down for the 11729C. Say your maximum PN level of interest is -50 dBc/Hz at 1 Hz. If you adjust the level going into the sound card to position the full, HPF'ed test signal near the sound card's rails, you'll have a theoretical floor near -170 dBc/Hz. Your signal generator might be set for -60 dB relative to the DUT input amplitude in this scenario, yielding a clean calibration spur without switching gains anywhere. The signal generator used to inject the beatnote for calibration doesn't need to be anything too fancy, just something with a good attenuator. > Many of the free applications that are out there will put a tone > out of the card and track it back into the card. That should at > least provide a tone to work with. I also should be something > that can be fairly easily verified. The issue of mixer output > impedance is still a little tricky without RF noise loading. It does need to be an RF tone, though, to characterize the whole signal path (and even then you'd really like to do it with broadband noise if it were practical, which it's not). If you haven't looked over HP note 11729B-1 ( http://www.ke5fx.com/gpib/5952-8286E.pdf ) it will give you a good refresher on what you're up against. -- john, KE5FX

Hi The only reason the 3561 takes in the beat note is because the front end gain changes automatically. Absolutely no different than switching an external preamp except the calibration sticker says not to worry about it. Yes indeed I have had that calibration sticker lie to me on occasion..... Bob On Feb 18, 2010, at 9:36 PM, John Miles wrote: > >> >> If the noise is "known flat" it's a good way to check system >> bandwidth. Some means of checking response is indeed very necessary. >> >> Not switching the preamp is indeed a good thing. The sound card >> does not have the range of the 3561, so with the sound card the >> beat note absolutely require a switch. The switch adds a second >> calibration step at audio. > > Maybe not -- the 3561A has nowhere near the dynamic range of the sound card > (13 bits versus 20+, or about 80 dB versus ~117 dB for a good sound card). > > Normally you do beatnote calibration by injecting a signal at (level of > DUT - gain of LNA), or 40 dB down for the 11729C. Say your maximum PN level > of interest is -50 dBc/Hz at 1 Hz. If you adjust the level going into the > sound card to position the full, HPF'ed test signal near the sound card's > rails, you'll have a theoretical floor near -170 dBc/Hz. Your signal > generator might be set for -60 dB relative to the DUT input amplitude in > this scenario, yielding a clean calibration spur without switching gains > anywhere. > > The signal generator used to inject the beatnote for calibration doesn't > need to be anything too fancy, just something with a good attenuator. > >> Many of the free applications that are out there will put a tone >> out of the card and track it back into the card. That should at >> least provide a tone to work with. I also should be something >> that can be fairly easily verified. The issue of mixer output >> impedance is still a little tricky without RF noise loading. > > It does need to be an RF tone, though, to characterize the whole signal path > (and even then you'd really like to do it with broadband noise if it were > practical, which it's not). If you haven't looked over HP note 11729B-1 ( > http://www.ke5fx.com/gpib/5952-8286E.pdf ) it will give you a good refresher > on what you're up against. > > -- john, KE5FX > > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. >

Bob Camp wrote: > If the noise is "known flat" it's a good way to check system bandwidth. Some means of checking response is indeed very necessary. > > Not switching the preamp is indeed a good thing. The sound card does not have the range of the 3561, so with the sound card the beat note absolutely require a switch. The switch adds a second calibration step at audio. > > To calibrate level, the levels of both the noise and carrier need to be well known. For a "basement system" the measure the slope as it crosses zero is likely more accurate. That of course assumes that you do the audio gain and response stuff properly. > > Many of the free applications that are out there will put a tone out of the card and track it back into the card. That should at least provide a tone to work with. I also should be something that can be fairly easily verified. The issue of mixer output impedance is still a little tricky without RF noise loading. One of those 346 A/B/C noise sources with 15 dB excess noise should be about right to check the noise floor. <http://freenet-homepage.de/dl4yhf/spectra1.html> looks like a good FFT program. regards, Gerhard

Bob Camp wrote: > Hi > > Ok, A bit more info: > > 1) Quadrature PLL using an RPD-1 DBM and a home brew lock box. > 2) Willingness to accept that I'm measuring a pair of oscillators > 3) Plenty of sources at the appropriate frequencies > 4) First took a shot at this in 1975 (I forget the Fluke app note number ...) been doing it ever since > 5) Appropriate preamp between the RPD and sound card is a work in progress I have built this preamp for this purpose: <http://www.hoffmann-hochfrequenz.de/downloads/preliminary_ssm2210.zip> Gerhard

Gerhard Hoffmann wrote: > Bob Camp wrote: >> Hi >> >> Ok, A bit more info: >> >> 1) Quadrature PLL using an RPD-1 DBM and a home brew lock box. 2) >> Willingness to accept that I'm measuring a pair of oscillators >> 3) Plenty of sources at the appropriate frequencies >> 4) First took a shot at this in 1975 (I forget the Fluke app note >> number ...) been doing it ever since >> 5) Appropriate preamp between the RPD and sound card is a work in >> progress > > I have built this preamp for this purpose: > > <http://www.hoffmann-hochfrequenz.de/downloads/preliminary_ssm2210.zip> > > Gerhard > Before deciding on a suitable preamp one should first measure the phase detector/mixer gain, noise noise and low frequency output impedance. The phase detector noise noise and gain depend on the phase detector output port termination network. For the common case of a diode ring phase detector where both inputs are saturated the output beat frequency waveform (and the phase detector output as a function of the phase difference between its input signals) for low beat frequencies will only be triangular if the phase detector output port termination impedance is low at the beat frequency and the sum frequency. If the termination network has a high impedance for low frequencies (<100kHz) and a low impedance for the sum frequency then the beat frequency output waveform will be approximately trapezoidal with rounded peaks. The range of impedances that can be considered low depends on the turns ratio of the transformers used in the mixer/phase detector. The mixer output impedance also depends on the transformer turns ratio. The phase detector output noise will be lower if the sum frequency is reflected back into the mixer than when the mixer is terminated in the specified resistive load. A preamp like that shown will be best suited to phase detectors with low gain and output impedance with a resistive output termination matched to the phase detector. Phase detectors like the Minicircuits RPD-1, MPD-1 etc have a relatively high output impedance and a preamp with a lower input current noise and somewhat higher input voltage noise may be better suited, particularly if the output is terminated in a capacitive input filter and the filter low frequency dc (and low frequency) load resistance is high. In such cases a version of the Wenzel JFET input amplifier with lower noise using fewer JFETs and having a low frequency input voltage noise that has a finite spectral density at dc may be more appropriate. An ultra low input offset voltage together with low drift is also possible. The only real requirement is that the preamp input noise (including the preamp input voltage noise and the noise voltage due to the preamp input noise current flowing in the phase detector output impedance ) spectral density should be significantly less (preferably at least 3x) than the phase detector output noise for all frequencies of interest. Bruce

Bruce Griffiths wrote: > > A preamp like that shown will be best suited to phase detectors with > low gain and output impedance with a resistive output termination > matched to the phase detector. > > Phase detectors like the Minicircuits RPD-1, MPD-1 etc have a > relatively high output impedance and a preamp with a lower input > current noise and somewhat higher input voltage noise may be better > suited, particularly if the output is terminated in a capacitive input > filter and the filter low frequency dc (and low frequency) load > resistance is high. > > In such cases a version of the Wenzel JFET input amplifier with lower > noise using fewer JFETs and having a low frequency input voltage noise > that has a finite spectral density at dc may be more appropriate. An > ultra low input offset voltage together with low drift is also possible. I have a version with BF862-FETs, too :-) Not bad, but close to the carrier, the BJTs win. BTW I don't like Wenzel's finesse noise remover any more, at least not for > 5 V. The large input coupling capacitor and fast raw power up/down can do bad things to the transistors BE diode, including driving it to zener for milliseconds. Gerhard

Gerhard Hoffmann wrote: > Bruce Griffiths wrote: >> >> A preamp like that shown will be best suited to phase detectors with >> low gain and output impedance with a resistive output termination >> matched to the phase detector. >> >> Phase detectors like the Minicircuits RPD-1, MPD-1 etc have a >> relatively high output impedance and a preamp with a lower input >> current noise and somewhat higher input voltage noise may be better >> suited, particularly if the output is terminated in a capacitive >> input filter and the filter low frequency dc (and low frequency) load >> resistance is high. >> >> In such cases a version of the Wenzel JFET input amplifier with lower >> noise using fewer JFETs and having a low frequency input voltage >> noise that has a finite spectral density at dc may be more >> appropriate. An ultra low input offset voltage together with low >> drift is also possible. > > I have a version with BF862-FETs, too :-) > Not bad, but close to the carrier, the BJTs win. > Not necessarily, it depends on the phase detector circuit details. What phase detector output termination network did you use? Aren't BF862's somewhat noisier than 2SK369's at low frequencies? > BTW I don't like Wenzel's finesse noise remover any more, at least not > for > 5 V. The large input coupling capacitor and fast raw power up/down > can do bad things to the transistors BE diode, including driving it to > zener > for milliseconds. > Easily fixed with a slight change to the circuit and a few diodes. It pays to simulate powering up circuits before building to many as its easier to fix such problems at that stage. Some low phase noise RF amplifier designs can even latch in a high current state if one isnt careful to fix such power up transient problems. > > Gerhard Bruce

Hi Don't forget that the output impedance of the mixer may also be changing with changes in input level. Lots of small things to check to make sure you have it all set up right. Bob On Feb 19, 2010, at 6:56 AM, Bruce Griffiths wrote: > Gerhard Hoffmann wrote: >> Bruce Griffiths wrote: >>> >>> A preamp like that shown will be best suited to phase detectors with low gain and output impedance with a resistive output termination matched to the phase detector. >>> >>> Phase detectors like the Minicircuits RPD-1, MPD-1 etc have a relatively high output impedance and a preamp with a lower input current noise and somewhat higher input voltage noise may be better suited, particularly if the output is terminated in a capacitive input filter and the filter low frequency dc (and low frequency) load resistance is high. >>> >>> In such cases a version of the Wenzel JFET input amplifier with lower noise using fewer JFETs and having a low frequency input voltage noise that has a finite spectral density at dc may be more appropriate. An ultra low input offset voltage together with low drift is also possible. >> >> I have a version with BF862-FETs, too :-) >> Not bad, but close to the carrier, the BJTs win. >> > Not necessarily, it depends on the phase detector circuit details. > What phase detector output termination network did you use? > Aren't BF862's somewhat noisier than 2SK369's at low frequencies? > > >> BTW I don't like Wenzel's finesse noise remover any more, at least not >> for > 5 V. The large input coupling capacitor and fast raw power up/down >> can do bad things to the transistors BE diode, including driving it to zener >> for milliseconds. >> > Easily fixed with a slight change to the circuit and a few diodes. > It pays to simulate powering up circuits before building to many as its easier to fix such problems at that stage. > Some low phase noise RF amplifier designs can even latch in a high current state if one isnt careful to fix such power up transient problems. >> >> Gerhard > Bruce > > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. >