Re: [time-nuts] About heterodyne method?
Sorry for the "for sale post" but if any one is interested I have a new to me Datum 1000B that seems to be a few hz high in frequency (after approx one week of warm up it would only adjust down to approx 10,000,001 Hz.) It might be usefull for this application. Please contact me off list if anyone has any interest in acquiring this unit. I'm looking for a few hundred plus shipping from Canada. I'm fairly flexible and will consider a try before you pay arrangment with longer term list members as long as the shipping fees are covered up front. I need to start clearing out some of my time nuts stuff (: Regards Mark Spencer --- On Fri, 7/27/12, Hui Zhang wrote: > From: Hui Zhang > Subject: Re: [time-nuts] About heterodyne method? > To: "Discussion of precise time and frequency measurement" > > Received: Friday, July 27, 2012, 9:28 AM > Hello Bob: > > Many thanks for your kindly and detailed > explanation, I think I am completely get it. I have few > minicircuit SBL-1 mixer module, but I lacks a limiter > circuit, I deicide to do some test near times. If I > have any progress I will be glad to tell you in this mail > list. > By use "DMTD" keyword, I got many useful > infomation of this method, when I finish original single > mixer circuit, I want try "DMTD" method late times. All of > these two method is interesting. > Thanks again for your help! > > > Hui > > > > > > At 2012-07-27 00:55:56,"Bob Camp" > wrote: > >Hi > > > >Quick summary: > > > >Computing / period counters give you a constant number > of digits of > >resolution regardless of input frequency. Back when this > stuff was developed > >a counter that gave you nine digits a second was pretty > common. > > > >Typical setup: > > > >1) Take two oscillators and tune them 1 to 10 Hz apart. > The technique only > >works if at least one of the oscillators can be tuned. > > > >2) Run the oscillators into a double balanced mixer. > Normally levels near > >the maximum are used on both inputs. > > > >3) Low pass filter the output. You want to keep the RF > out of the 1 to 10 Hz > >beat note. Various terminations seem to help the > sensitivity of various > >mixers (high impedance at audio, reactive termination at > RF etc). > > > >4) Run the audio beat note into a limiter. The design of > the limiter can be > >fairly simple or quite elaborate. You need the limiter > because the counters > >input channels rarely do well with audio sine waves. > > > >5) Count the frequency of the audio on your counter. > > > >If you start off with 10 MHz and set them 1 Hz apart, > you get a 1x10^7 > >"amplification" of the frequency error. If you count > that to nine digits, > >the resulting resolution would be 1x10^-15. > > > >Of course resolution and accuracy are not the same > thing. Measurement noise > >will dominate the readings past a certain point. A > typical setup should get > >you below 1x10^-11 at one second without a lot of > effort. A good setup can > >get you to 5x10^-13 at one second with common parts. A > fancy setup with a > >complex limiter can go below 1x10^-13 at one second. > > > >Yes indeed there's a bit more than that to it, but that > should get you > >started. > > > >Bob > > > >-Original Message- > >From: time-nuts-boun...@febo.com > [mailto:time-nuts-boun...@febo.com] > On > >Behalf Of Hui Zhang > >Sent: Thursday, July 26, 2012 10:44 AM > >To: time-nuts@febo.com > >Subject: [time-nuts] About heterodyne method? > > > >Hello everyone: > > > > I hnow 'heterodyne method' is very useful > method of pricision frequency > >measurement (use DBM and LPF), but I only found very few > infomation when I > >searched in early docments, Can someone give me > more information about this > >'heterodyne method'? The block diagram is best. I will > very appreciate of > >that. > > > > > >Hui > >___ > >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.
Re: [time-nuts] About heterodyne method?
Hello Bob: Many thanks for your kindly and detailed explanation, I think I am completely get it. I have few minicircuit SBL-1 mixer module, but I lacks a limiter circuit, I deicide to do some test near times. If I have any progress I will be glad to tell you in this mail list. By use "DMTD" keyword, I got many useful infomation of this method, when I finish original single mixer circuit, I want try "DMTD" method late times. All of these two method is interesting. Thanks again for your help! Hui At 2012-07-27 00:55:56,"Bob Camp" wrote: >Hi > >Quick summary: > >Computing / period counters give you a constant number of digits of >resolution regardless of input frequency. Back when this stuff was developed >a counter that gave you nine digits a second was pretty common. > >Typical setup: > >1) Take two oscillators and tune them 1 to 10 Hz apart. The technique only >works if at least one of the oscillators can be tuned. > >2) Run the oscillators into a double balanced mixer. Normally levels near >the maximum are used on both inputs. > >3) Low pass filter the output. You want to keep the RF out of the 1 to 10 Hz >beat note. Various terminations seem to help the sensitivity of various >mixers (high impedance at audio, reactive termination at RF etc). > >4) Run the audio beat note into a limiter. The design of the limiter can be >fairly simple or quite elaborate. You need the limiter because the counters >input channels rarely do well with audio sine waves. > >5) Count the frequency of the audio on your counter. > >If you start off with 10 MHz and set them 1 Hz apart, you get a 1x10^7 >"amplification" of the frequency error. If you count that to nine digits, >the resulting resolution would be 1x10^-15. > >Of course resolution and accuracy are not the same thing. Measurement noise >will dominate the readings past a certain point. A typical setup should get >you below 1x10^-11 at one second without a lot of effort. A good setup can >get you to 5x10^-13 at one second with common parts. A fancy setup with a >complex limiter can go below 1x10^-13 at one second. > >Yes indeed there's a bit more than that to it, but that should get you >started. > >Bob > >-Original Message- >From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On >Behalf Of Hui Zhang >Sent: Thursday, July 26, 2012 10:44 AM >To: time-nuts@febo.com >Subject: [time-nuts] About heterodyne method? > >Hello everyone: > >I hnow 'heterodyne method' is very useful method of pricision frequency >measurement (use DBM and LPF), but I only found very few infomation when I >searched in early docments, Can someone give me more information about this >'heterodyne method'? The block diagram is best. I will very appreciate of >that. > > >Hui >___ >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.
Re: [time-nuts] About heterodyne method?
Hello Azelio: Thank you very much, I use this keyword searched in internet and got many useful infomation, include some block diagram. Thank you! Hui At 2012-07-26 22:55:26,"Azelio Boriani" wrote: >Search for DMTD, dual mixer time difference. > >On Thu, Jul 26, 2012 at 4:43 PM, Hui Zhang wrote: > >> Hello everyone: >> >> I hnow 'heterodyne method' is very useful method of pricision >> frequency measurement (use DBM and LPF), but I only found very few >> infomation when I searched in early docments, Can someone give me more >> information about this 'heterodyne method'? The block diagram is best. I >> will very appreciate of that. >> >> >> Hui >> ___ >> 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.
Re: [time-nuts] About heterodyne method?
Hi Quick summary: Computing / period counters give you a constant number of digits of resolution regardless of input frequency. Back when this stuff was developed a counter that gave you nine digits a second was pretty common. Typical setup: 1) Take two oscillators and tune them 1 to 10 Hz apart. The technique only works if at least one of the oscillators can be tuned. 2) Run the oscillators into a double balanced mixer. Normally levels near the maximum are used on both inputs. 3) Low pass filter the output. You want to keep the RF out of the 1 to 10 Hz beat note. Various terminations seem to help the sensitivity of various mixers (high impedance at audio, reactive termination at RF etc). 4) Run the audio beat note into a limiter. The design of the limiter can be fairly simple or quite elaborate. You need the limiter because the counters input channels rarely do well with audio sine waves. 5) Count the frequency of the audio on your counter. If you start off with 10 MHz and set them 1 Hz apart, you get a 1x10^7 "amplification" of the frequency error. If you count that to nine digits, the resulting resolution would be 1x10^-15. Of course resolution and accuracy are not the same thing. Measurement noise will dominate the readings past a certain point. A typical setup should get you below 1x10^-11 at one second without a lot of effort. A good setup can get you to 5x10^-13 at one second with common parts. A fancy setup with a complex limiter can go below 1x10^-13 at one second. Yes indeed there's a bit more than that to it, but that should get you started. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Hui Zhang Sent: Thursday, July 26, 2012 10:44 AM To: time-nuts@febo.com Subject: [time-nuts] About heterodyne method? Hello everyone: I hnow 'heterodyne method' is very useful method of pricision frequency measurement (use DBM and LPF), but I only found very few infomation when I searched in early docments, Can someone give me more information about this 'heterodyne method'? The block diagram is best. I will very appreciate of that. Hui ___ 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.
Re: [time-nuts] About heterodyne method?
Search for DMTD, dual mixer time difference. On Thu, Jul 26, 2012 at 4:43 PM, Hui Zhang wrote: > Hello everyone: > > I hnow 'heterodyne method' is very useful method of pricision > frequency measurement (use DBM and LPF), but I only found very few > infomation when I searched in early docments, Can someone give me more > information about this 'heterodyne method'? The block diagram is best. I > will very appreciate of that. > > > Hui > ___ > 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.
