From: Cube Central <cubecentral@gmail.com>

Yeah, I just have it plugged into my LAN.  Timelab finds it when I start it up.  It's a bit more complicated to find it from a C program, but there are instructions to help you in their documentation.  And ethernet addresses are a lot easier than the ring-around-the-rosie that can happen with USB addresses.  I think you'll like it. Bob ----------------------------------------------------------------- AE6RV.com GFS GPSDO list: groups.yahoo.com/neo/groups/GFS-GPSDOs/info From: Cube Central <cubecentral@gmail.com> To: 'Discussion of precise time and frequency measurement' <time-nuts@febo.com> Sent: Friday, October 7, 2016 2:03 PM Subject: Re: [time-nuts] Inexpensive Alternative to a 5120A Thanks again, Bob.  I've just ordered the controller you suggested.  No wonder I was running into a mental block, as the port on the back of the 5335A counter says "HP IB" and there was all this talk of "GPIB" and I simply didn't put 2 and 2 together. I gather that the Ethernet one will work over a standard LAN and that the Timelab software will have no issue finding it (once configured?)  I should - and am about to - start reading on how to set up that controller.  I was thinking that the USB one might be more direct into a neighboring computer, but not knowing anything about it, will go with your suggestion. This certainly opens up many new avenues for me to continue learning as I am ready for this next step to bring a PC to bear on measurements.  I am glad that I had the accidental foresight to get the correct TIC!  Thanks again and any further suggestions on where to start (especially what to plug into what as described below which I found helpful) is greatly appreciated.      -Randal R.         (at CubeCentral) -----Original Message----- From: time-nuts [mailto:time-nuts-bounces@febo.com] On Behalf Of Bob Stewart Sent: Friday, 07 October, 2016 12:43 To: Discussion of precise time and frequency measurement <time-nuts@febo.com> Subject: Re: [time-nuts] Inexpensive Alternative to a 5120A Here's the one I recommend:  http://prologix.biz/gpib-ethernet-controller.html No, it's not a $35 adapter, but IMO it's worth every dime because 1) it's ethernet and not USB and 2) I can write C programs to interface it without knowing anything about the underlying GPIB protocol.  Also, Timelab talks to it just fine.  (Disclaimer:  I have no financial or other relationship with the Prologix people.) For your timing tests, you'll probably want to connect a cable from the 10MHz TIME BASE out in the back to channel B in the front.  Send the 1PPS to Channel A and use Function Time A->B.  To measure two 10MHz signals, connect them to A and B and route the 1PPS signal to the arming input on the back. Bob ----------------------------------------------------------------- AE6RV.com GFS GPSDO list: groups.yahoo.com/neo/groups/GFS-GPSDOs/info       From: Cube Central <cubecentral@gmail.com> To: 'Discussion of precise time and frequency measurement' <time-nuts@febo.com> Sent: Friday, October 7, 2016 1:32 PM Subject: Re: [time-nuts] Inexpensive Alternative to a 5120A   Thanks Bob, I hope your cold improves, and I thank you for that answer.  As I happen to have a HP 5335A and I think it has the options 10 and 30.  So from there, you are saying that the next step would be a GPIB adapter.  Here be dragons, as I have never used anything with GPIB before.  I am happy to order one today if you (or someone) could point to exactly the one I should get (perhaps an e-bay link?) or are all these pretty much standard? I also have a very nice GPSDO thanks to Nick Sayer's project:  https://hackaday.io/project/6872-gps-disciplined-xcxo I assume that things will become clear the more I research GPIB and adapters?  I think that is the missing piece I was looking for in order to really use the Timelab software. Thanks again for the help, all you time-nuts, as I seem to be ready to take the next step...     -Randal R.         (at CubeCentral) -----Original Message----- From: time-nuts [mailto:time-nuts-bounces@febo.com] On Behalf Of Bob Stewart Sent: Friday, 07 October, 2016 11:21 To: Discussion of precise time and frequency measurement <time-nuts@febo.com> Subject: Re: [time-nuts] Inexpensive Alternative to a 5120A Hi Randal, I've got a bad cold, so I can only hope this response is rational. To begin with, you need some sort of Time Interval Counter (TIC), as well as some way to get the measurements from it into your computer.  Once you get that, then you get a free copy of John Miles' indispensable Timelab software and have at it. A beginner setup might consist of an HP 5335A, which can be found cheaply, and a GPIB adapter.  Be sure to get a 5335A with the high stability OCXO; i.e. the venerable HP 10811.  You can find both on ebay, but the usual cautions apply.  Personally, I prefer the Prologix Ethernet GPIB adapter.  It's pricey at about $200, but it's the easiest for mere mortals to interface to with custom programs that you may be tempted to write. If you want to get a bit better than that, then you'll probably go with an HP 5370A with a 10811.  Other companies make good TICs, as well.  At some point you'll probably want to get a better standard than the 10811 in your counter, so you'll consider either a Rubidium standard or a Cesium standard.  Personally, I'd avoid the FE-5680s on ebay, as they're not as stable as you'd expect them to be.  But, other people may disagree.  Cesium standards aren't cheap - unless you happen to make the score of a lifetime as another time-nut did with a PRS-45A. You can also use a GPSDO as a reference.  I sell a new one, and you can find good surplus units on ebay at various prices.  Different GPSDOs will have different qualities.  They all suffer from to extent from ionospheric drift.  Older units with older receivers probably have the most, but I'm not an expert on GPSDO selection. So, with all that said, start with the cheapest you can afford, and then decide whether you really want to get into serious time-nuttery.  It's not an inexpensive hobby, even at the low end.  But it's probably cheaper than a diet of loose women and beer! Bob - AE6RV ----------------------------------------------------------------- AE6RV.com GFS GPSDO list: groups.yahoo.com/neo/groups/GFS-GPSDOs/info       From: Cube Central <cubecentral@gmail.com> To: 'Discussion of precise time and frequency measurement' <time-nuts@febo.com> Sent: Friday, October 7, 2016 11:52 AM Subject: [time-nuts] Inexpensive Alternative to a 5120A   I saw that someone was using one of these: http://www.microsemi.com/products/timing-synchronization-systems/test-measurement/test-sets/5120a ...and the more I looked at it, the more I wanted one.  However, I couldn't get a price for one, but only quotes.  This can only mean that they are really expensive and out of reach of mere mortal novice time-nut like myself. Is there an alternative that someone could point me to that would cost only a couple hundred rather than (what I expect) is a couple thousand?  How would I go about gathering the data needed for these nifty ADEV graphs I see floating about in here? Could there be a kit or something that I can plug into a PC?  I really don't even know where to start looking? Thanks - in advance - for any suggestions!     -Randal R.         (at CubeCentral) _______________________________________________ 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.

On 10/7/16 12:03 PM, Cube Central wrote: > Thanks again, Bob. I've just ordered the controller you suggested. > No wonder I was running into a mental block, as the port on the back > of the 5335A counter says "HP IB" and there was all this talk of > "GPIB" and I simply didn't put 2 and 2 together. > GPIB == HPIB == IEEE-488 HP invented it, so they, of course, call it HPIB. > I gather that the Ethernet one will work over a standard LAN and that > the Timelab software will have no issue finding it (once configured?) > I should - and am about to - start reading on how to set up that > controller. I was thinking that the USB one might be more direct > into a neighboring computer, but not knowing anything about it, will > go with your suggestion. It's easy to configure.. I give it a static IP (e.g. 192.168.1.100), and then you're off to the races.. > > This certainly opens up many new avenues for me to continue learning > as I am ready for this next step to bring a PC to bear on > measurements. I am glad that I had the accidental foresight to get > the correct TIC! Thanks again and any further suggestions on where > to start (especially what to plug into what as described below which > I found helpful) is greatly appreciated. > > -Randal R. (at CubeCentral) > > -----Original Message----- From: time-nuts > [mailto:time-nuts-bounces@febo.com] On Behalf Of Bob Stewart Sent: > Friday, 07 October, 2016 12:43 To: Discussion of precise time and > frequency measurement <time-nuts@febo.com> Subject: Re: [time-nuts] > Inexpensive Alternative to a 5120A > > Here's the one I recommend: > http://prologix.biz/gpib-ethernet-controller.html > > No, it's not a $35 adapter, but IMO it's worth every dime because 1) > it's ethernet and not USB and 2) I can write C programs to interface > it without knowing anything about the underlying GPIB protocol. > Also, Timelab talks to it just fine. (Disclaimer: I have no > financial or other relationship with the Prologix people.) > > For your timing tests, you'll probably want to connect a cable from > the 10MHz TIME BASE out in the back to channel B in the front. Send > the 1PPS to Channel A and use Function Time A->B. To measure two > 10MHz signals, connect them to A and B and route the 1PPS signal to > the arming input on the back. > > Bob > ----------------------------------------------------------------- > AE6RV.com > > GFS GPSDO list: groups.yahoo.com/neo/groups/GFS-GPSDOs/info > > From: Cube Central <cubecentral@gmail.com> To: 'Discussion of precise > time and frequency measurement' <time-nuts@febo.com> Sent: Friday, > October 7, 2016 1:32 PM Subject: Re: [time-nuts] Inexpensive > Alternative to a 5120A > > Thanks Bob, I hope your cold improves, and I thank you for that > answer. > > As I happen to have a HP 5335A and I think it has the options 10 and > 30. So from there, you are saying that the next step would be a GPIB > adapter. Here be dragons, as I have never used anything with GPIB > before. I am happy to order one today if you (or someone) could > point to exactly the one I should get (perhaps an e-bay link?) or are > all these pretty much standard? > > I also have a very nice GPSDO thanks to Nick Sayer's project: > https://hackaday.io/project/6872-gps-disciplined-xcxo > > I assume that things will become clear the more I research GPIB and > adapters? I think that is the missing piece I was looking for in > order to really use the Timelab software. > > Thanks again for the help, all you time-nuts, as I seem to be ready > to take the next step... > > -Randal R. (at CubeCentral) > > > -----Original Message----- From: time-nuts > [mailto:time-nuts-bounces@febo.com] On Behalf Of Bob Stewart Sent: > Friday, 07 October, 2016 11:21 To: Discussion of precise time and > frequency measurement <time-nuts@febo.com> Subject: Re: [time-nuts] > Inexpensive Alternative to a 5120A > > Hi Randal, I've got a bad cold, so I can only hope this response is > rational. To begin with, you need some sort of Time Interval Counter > (TIC), as well as some way to get the measurements from it into your > computer. Once you get that, then you get a free copy of John Miles' > indispensable Timelab software and have at it. A beginner setup might > consist of an HP 5335A, which can be found cheaply, and a GPIB > adapter. Be sure to get a 5335A with the high stability OCXO; i.e. > the venerable HP 10811. You can find both on ebay, but the usual > cautions apply. Personally, I prefer the Prologix Ethernet GPIB > adapter. It's pricey at about $200, but it's the easiest for mere > mortals to interface to with custom programs that you may be tempted > to write. If you want to get a bit better than that, then you'll > probably go with an HP 5370A with a 10811. Other companies make good > TICs, as well. At some point you'll probably want to get a better > standard than the 10811 in your counter, so you'll consider either a > Rubidium standard or a Cesium standard. Personally, I'd avoid the > FE-5680s on ebay, as they're not as stable as you'd expect them to > be. But, other people may disagree. Cesium standards aren't cheap - > unless you happen to make the score of a lifetime as another time-nut > did with a PRS-45A. You can also use a GPSDO as a reference. I sell > a new one, and you can find good surplus units on ebay at various > prices. Different GPSDOs will have different qualities. They all > suffer from to extent from ionospheric drift. Older units with older > receivers probably have the most, but I'm not an expert on GPSDO > selection. So, with all that said, start with the cheapest you can > afford, and then decide whether you really want to get into serious > time-nuttery. It's not an inexpensive hobby, even at the low end. > But it's probably cheaper than a diet of loose women and beer! > > Bob - AE6RV > ----------------------------------------------------------------- > AE6RV.com > > GFS GPSDO list: groups.yahoo.com/neo/groups/GFS-GPSDOs/info > > From: Cube Central <cubecentral@gmail.com> To: 'Discussion of precise > time and frequency measurement' <time-nuts@febo.com> Sent: Friday, > October 7, 2016 11:52 AM Subject: [time-nuts] Inexpensive Alternative > to a 5120A > > I saw that someone was using one of these: > http://www.microsemi.com/products/timing-synchronization-systems/test-measurement/test-sets/5120a > > ...and the more I looked at it, the more I wanted one. However, I > couldn't get a price for one, but only quotes. This can only mean > that they are really expensive and out of reach of mere mortal novice > time-nut like myself. > > Is there an alternative that someone could point me to that would > cost only a couple hundred rather than (what I expect) is a couple > thousand? How would I go about gathering the data needed for these > nifty ADEV graphs I see floating about in here? > > Could there be a kit or something that I can plug into a PC? I > really don't even know where to start looking? > > Thanks - in advance - for any suggestions! > > -Randal R. (at CubeCentral) > > > > _______________________________________________ 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 Randal, On 2016-10-07 18:52:57, Cube Central wrote: > Is there an alternative that someone could point me to that would cost only a couple hundred rather than (what I expect) is a couple thousand? How would I go about gathering the data needed for these nifty ADEV graphs I see floating about in here? People have reported (also on this list) that some SDR (software defined radio) hardware is quite capable of emulating a 5120/5125, and even going beyond it in performance. "Oscillator metrology with software defined radio, Sherman, J.A., Jördens, R." - arXiv:1605.03505 [physics.ins-det] https://arxiv.org/abs/1605.03505 Regards, Paul Boven.

Likely the lowest cost way to get into that is with a TV tuner USB dongle. They cost about $20. People are able to get about 2.4 mega samples per second. Not a lot of dynamic range but you can control that. Use a mixer to move the signal of interest into the range the tuner can handle. Tuniers typically tune from about 20Mhz to 1Ghz or 2Ghz approximate. On Sun, Oct 9, 2016 at 12:05 AM, Paul Boven <p.boven@xs4all.nl> wrote: > Hi Randal, > > On 2016-10-07 18:52:57, Cube Central wrote: >> >> Is there an alternative that someone could point me to that would cost >> only a couple hundred rather than (what I expect) is a couple thousand? How >> would I go about gathering the data needed for these nifty ADEV graphs I see >> floating about in here? > > > People have reported (also on this list) that some SDR (software defined > radio) hardware is quite capable of emulating a 5120/5125, and even going > beyond it in performance. > > "Oscillator metrology with software defined radio, Sherman, J.A., Jördens, > R." - arXiv:1605.03505 [physics.ins-det] > https://arxiv.org/abs/1605.03505 > > Regards, Paul Boven. > > > _______________________________________________ > 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. -- Chris Albertson Redondo Beach, California

Hi > On Oct 9, 2016, at 6:18 PM, Chris Albertson <albertson.chris@gmail.com> wrote: > > Likely the lowest cost way to get into that is with a TV tuner USB > dongle. They cost about $20. People are able to get about 2.4 mega > samples per second. Except that you need about 30 mega samples ... > Not a lot of dynamic range but you can control > that. Use a mixer to move the signal of interest into the range > the tuner can handle. Tuniers typically tune from about 20Mhz to > 1Ghz or 2Ghz approximate. You also need a very specific dual ADC architecture as described in the paper. Their hardware was > $1K and probably the only suitable system at that low a price. No free lunch :) Bob > > > On Sun, Oct 9, 2016 at 12:05 AM, Paul Boven <p.boven@xs4all.nl> wrote: >> Hi Randal, >> >> On 2016-10-07 18:52:57, Cube Central wrote: >>> >>> Is there an alternative that someone could point me to that would cost >>> only a couple hundred rather than (what I expect) is a couple thousand? How >>> would I go about gathering the data needed for these nifty ADEV graphs I see >>> floating about in here? >> >> >> People have reported (also on this list) that some SDR (software defined >> radio) hardware is quite capable of emulating a 5120/5125, and even going >> beyond it in performance. >> >> "Oscillator metrology with software defined radio, Sherman, J.A., Jördens, >> R." - arXiv:1605.03505 [physics.ins-det] >> https://arxiv.org/abs/1605.03505 >> >> Regards, Paul Boven. >> >> >> _______________________________________________ >> 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. > > > > -- > > Chris Albertson > Redondo Beach, California > _______________________________________________ > 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.

The Red Pitaya is a lower cost (but lower performance) alternative to the USRP boxes. I've implemented a RF phase meter using the RP and get about 1 ps time resolution at an averaging time of 1 s. The RP analog input side seems to be a bit noisy so loses a few bits. There was some discussion about this on timenuts some time back. Someone has implemented an SDR on the RP http://redpitaya.com/red-pitaya-and-software-defined-radio/ - this may be useable in the same way as described in the NIST paper. It's on my list to try it out some day. The summary is that if you only want to pay $400, rather than $40 000, you have to do some/a lot of work. Cheers Michael On Mon, Oct 10, 2016 at 10:16 AM, Bob Camp <kb8tq@n1k.org> wrote: > Hi > > >> On Oct 9, 2016, at 6:18 PM, Chris Albertson <albertson.chris@gmail.com> wrote: >> >> Likely the lowest cost way to get into that is with a TV tuner USB >> dongle. They cost about $20. People are able to get about 2.4 mega >> samples per second. > > Except that you need about 30 mega samples ... > > >> Not a lot of dynamic range but you can control >> that. Use a mixer to move the signal of interest into the range >> the tuner can handle. Tuniers typically tune from about 20Mhz to >> 1Ghz or 2Ghz approximate. > > You also need a very specific dual ADC architecture as described in the paper. Their > hardware was > $1K and probably the only suitable system at that low a price. > > No free lunch :) > > Bob > >> >> >> On Sun, Oct 9, 2016 at 12:05 AM, Paul Boven <p.boven@xs4all.nl> wrote: >>> Hi Randal, >>> >>> On 2016-10-07 18:52:57, Cube Central wrote: >>>> >>>> Is there an alternative that someone could point me to that would cost >>>> only a couple hundred rather than (what I expect) is a couple thousand? How >>>> would I go about gathering the data needed for these nifty ADEV graphs I see >>>> floating about in here? >>> >>> >>> People have reported (also on this list) that some SDR (software defined >>> radio) hardware is quite capable of emulating a 5120/5125, and even going >>> beyond it in performance. >>> >>> "Oscillator metrology with software defined radio, Sherman, J.A., Jördens, >>> R." - arXiv:1605.03505 [physics.ins-det] >>> https://arxiv.org/abs/1605.03505 >>> >>> Regards, Paul Boven. >>> >>> >>> _______________________________________________ >>> 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. >> >> >> >> -- >> >> Chris Albertson >> Redondo Beach, California >> _______________________________________________ >> 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.

Let's back up, start over. I want to use an SDR-like thing to characterize my new 10MHz oscillator, or actually to compare it to my Thunderbolt What are the steps? Then after we list the steps I want to see why so many samples are required Or maybe a better way to ask is "How will the results be limited if I can only sample at 2.3 MSPS? Obviously I'n not directly sampling there 10MHz signal. I'm sampling there frequency difference that a mixer gives me. On Sun, Oct 9, 2016 at 4:16 PM, Bob Camp <kb8tq@n1k.org> wrote: > Hi > > >> On Oct 9, 2016, at 6:18 PM, Chris Albertson <albertson.chris@gmail.com> wrote: >> >> Likely the lowest cost way to get into that is with a TV tuner USB >> dongle. They cost about $20. People are able to get about 2.4 mega >> samples per second. > > Except that you need about 30 mega samples ... -- Chris Albertson Redondo Beach, California

Hi > On Oct 10, 2016, at 7:06 AM, Chris Albertson <albertson.chris@gmail.com> wrote: > > Let's back up, start over. I want to use an SDR-like thing to > characterize my new 10MHz oscillator, or actually to compare it to my > Thunderbolt > What are the steps? The test document on an OCXO or a GPSDO runs into several dozen to several hundred pages. It also sub-references multiple documents of similar size. All that gets only just so deep. The assumption is that routine maintenance and calibration have taken care of a lot of other things. You do *not* want to dig into all that :) > > Then after we list the steps I want to see why so many samples are required The original paper did a direct sample of the signal. Nyquist gets you into the > 20 mega sample range. > > Or maybe a better way to ask is "How will the results be limited if I > can only sample at 2.3 MSPS? They use the dual ADC simultaneous sample process to reduce the jitter of the sample. It’s in the paper. > > Obviously I'n not directly sampling there 10MHz signal. I'm sampling > there frequency difference that a mixer gives me. Then build a DMTD with a stable / low noise reference rather than the poor stability reference in a SDR dongle. Again, no free lunch :) Bob > > > On Sun, Oct 9, 2016 at 4:16 PM, Bob Camp <kb8tq@n1k.org> wrote: >> Hi >> >> >>> On Oct 9, 2016, at 6:18 PM, Chris Albertson <albertson.chris@gmail.com> wrote: >>> >>> Likely the lowest cost way to get into that is with a TV tuner USB >>> dongle. They cost about $20. People are able to get about 2.4 mega >>> samples per second. >> >> Except that you need about 30 mega samples ... > -- > > Chris Albertson > Redondo Beach, California > _______________________________________________ > 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.