Re: [time-nuts] Good references on holdover?
You may find the following Master's thesis useful: http://www.ko4bb.com/manuals/download.php?file=08_Stuff_Not_Sorted/8_Sept_28_2014_Uploads/Adaptive_OCXO_drift_correction_thesis_Zhou_2009.pdf Best regards, Charles ___ 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] Racal-Dana GPIB card question
You might want to join the Racal-Dana Yahoo group for help on this. There is a gotcha with these cards in that S4 selects which 488 protocol is used( Airforce or normal). Apparently you just have to swap it over and re-power. Sorry cannot help you with your software level matching question. Regards Malcolm ___ 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] T.I. questions
Several of the reciprocal counters (DC509, DC5010) Tektronix built for their TM500/TM5000 test equipment mainframes use a National Semiconductor noise generator chip to dither their reference clock. They do this mainly to handle the case where the input freq and reference clock are very close. The noise generator chip is not a true noise generator. I think it used a ROM table to play back a fixed random bit pattern. ___ 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] HP 5065A questions
cdelect@... writes: Joe, Nice find. Don't worry about the lamp, they VERY seldom fail. Of course there are electronics failures that crop up. What color is your physics package? Blue paint= old style Olive green paint = mid production Silver (no paint) = late production Also can I get your units serial number for my list? With it in the operate mode watch the control voltage and slowly adjust the fine quartz back and forth. If it is indeed locking to the Rubidium the meter should follow your adjustment. I would not worry about the physics package case either, just put some tape across to seal it from heat loss. I can give you some alignment hints if you need them. Corby Dawson cdelect@... The physics package is silver so you reckon late production. By looking at date codes I reckon it's late 1988 production, does that come into that definition. When did they stop making the unit? Also regarding serial number the tag at the back has been filed so that the number is not legible, however the calibration stickers and a departmental identification tag list it as 916-00184, does that tie in with the rest of the numbers you have? Again from the cal stickers etc this is an ex RAF machine. I switched to control voltage and initially it was off the scale, adjusting the fine control just brought it back to 50, so I tweaked the coarse slightly to take the reading down to 40 and then adjusted the fine as you suggest and the meter follows the direction of adjustment. This also took the output from +25 millihertz to -2 around the 5MHz nominal so it does appear to be locked. The unit does not have any options so just a blank plate on the right and the lock light is red instead of green also the back panel connections are different to any shown in the manuals I've seen, though I have found some of the differences by trawling through the changes section. I'm back to the cal lab next week to pick up a meter that they are calling for me and whilst there I shall see if I can find something of the same vintage that uses the same top and bottom covers and side rails that I can cannibalize to fix this unit. If I can then I'll keep and get it back to pristine condition, if I can't then I'll find someone else who want's to take on the burden. JoeD ___ 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] T.I. questions
The typical noise generator chips uses a PRNG based on DFFs and XOR gate(s). A typical weakness is that the chain of DFFs is to short, causing a relatively high rate of cycling, which hearable as a beating. However, for some uses, that is OK. Cheers, Magnus On 02/06/2015 07:16 PM, Mark Sims wrote: Several of the reciprocal counters (DC509, DC5010) Tektronix built for their TM500/TM5000 test equipment mainframes use a National Semiconductor noise generator chip to dither their reference clock. They do this mainly to handle the case where the input freq and reference clock are very close. The noise generator chip is not a true noise generator. I think it used a ROM table to play back a fixed random bit pattern. ___ 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] T.I. questions
mag...@rubidium.dyndns.org said: The typical noise generator chips uses a PRNG based on DFFs and XOR gate(s). A typical weakness is that the chain of DFFs is to short, causing a relatively high rate of cycling, which hearable as a beating. However, for some uses, that is OK. The buzzword for the typical PRNG is LFSR - Linear Feedback Shift Register. Many years ago, Xilinx published a good app-note on this topic. There is also a section in Art of Electronics. With the right generating polynomial, you get a sequence of bits that doesn't repeat until 2^N-1 bits. The math geeks like to collect them. Each 1 bit in the polynomial turns into an XOR gate, so you will also find collections of polynomials with fewest bits. It's hard to imagine serious problems with too-short. Each FF doubles the no-repeat length. -- These are my opinions. I hate spam. ___ 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] 1 PPS Correction Problem using DS1123LE-50, Delay Line
Could it maybe be that you're actually writing the delay value into the chip while the pulse is high? I've had that problem with a DS1023-100. The solution was to wait until the pulse goes low, and then set the delay for the next pulse. Regards, Tom On Wed, Feb 4, 2015 at 2:47 PM, Dan Kemppainen d...@irtelemetrics.com wrote: Martyn, In working with a 6T receiver we have noticed that sometimes the reported offset correction is not correct. These instances appear to be infrequent, and have only been noted as a single instance at a time. For example you will occasionally get a -10.2nS correction when you should get a positive one, or vice versa. Simply changing the sign of the offending correction value appears to make it correct. This doesn't sound exactly like what you are seeing, however. Hope the info helps! Dan On 2/4/2015 12:00 PM, time-nuts-requ...@febo.com wrote: Hello, I am using a LEA-6T GPS chip with the DS1123LE-50 delay line to correct the jitter of the 1 pps output from the GPS chip. The correction works very well, dropping the peak to peak jitter of the 1pps output from about 20 ns to 3 ns. However, I randomly get a big jump of the 1 pps pulse, by many millisceconds. After lots of research, checking code and hardware, it seems the DS1123LE-50 chip randomly produces it's own long pulse. I have monitored the pulse going into and out of the delay line The input pulse remains constant (I can't remember exactly the pulse width, I think 10 us). But suddenly a large output pulse of many milliseconds appear. Some chips don't seem to have this problem. Other chips can go wrong, 20 times in an hour, then run perfectly for a day or so. I am blaming my software guy, he is blaming the chip. Has anyone else used this delay line and seen anything similar? This chip is now apparently obsolete, but I don't want to change yet as I have a few years worth of stock of these chips. Regards Martyn ___ 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] D term (was no subject)
In message 20150206153214.4d5f42edbdda4639fee1a...@kinali.ch, Attila Kinali w rites: On Mon, 26 Jan 2015 20:15:12 + Poul-Henning Kamp p...@phk.freebsd.dk wrote: The basic math of PID has been around for about 100 years. The invention of the servo (and synchro/resolver) is what makes its day... If anyone wants to dive into control theory I recommend reading the book Feedback control of dynamic systems by Franklin, Powell and Emami-Naeini. And if you are more of a historical bent, the MIT Radiation Lab series is the motherlode. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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] Newbie question
Hi Time-Nuts: Not sure what the protocol is here but I'll just jump in. I've just purchased an HP53310a modulation domain analyzer. Most you already know that these amazing instruments are basically a TIC with a graphic display of frequency vs time. I've always wanted one to record PLL settling time. I also know they are pretty non-intuitive to setup and use. Now that I have one I'd like to connect with someone that has experience using them. Anyone? Best Regards to the group. Stuart Rumley 650-369-0575 Hi Stuart, Also contact Joe Geller, who was active on time-nuts some years ago, and who wrote the ultimate 53310a reference page: http://www.gellerlabs.com/hp53310a.htm /tvb One thing that Joe Geller doesn't mention is the fierce leakage from the power transformer. You may not notice it on a modern scope with LCD display, but my 53310A spewed so much field that it demolished the display on a Tek 2467B sitting nearby. I had to move the 53310A to a separate bench. Why it didn't seem to bother its own internal crt is a complete mystery. I'm sure it would bother a Rubidium. I found the 53310A to be a lot less useful than expected. There are other ways of making measurements that are easier and more accurate. Now, it just sits in a corner gathering dust. As far as measuring the PLL settling, you will need an external reference that you can frequency modulate. Depending on your requirements, it could be something simple like an RC oscillator with lots of phase noise, or two low phase noise OCXO's at slightly different frequencies with a digital mixer detecting the phase alignment and switching the signal to your PLL from one OCXO to the other. You can monitor the VCO DC error voltage to look for risetime and ringing problems. To monitor settling time, trigger the scope on the frequency switch, delay out to the desired region, and display the reference and VCO signals to the phase detector added together. If you use a triggered delay you can step from one cycle to the next and watch the signals align themselves. Depending on the skill of the designer, the PLL may exhibit jumps in phase during settling due to crosstalk, or many other kinds of nasties at different points in the phase relationship. If he did not monitor the PLL settling during development, there will likely be things you may not like. Bottom line is I much prefer looking at the VCO DC error and the inputs to the phase detector rather than trying to investigate these problems with the 53310A. The actual waveforms give a lot more information to work with. Mike Monett ___ 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] HP 5065A questions
Joe a very nice find. The light won't change to lock unless you toggle the little switch inside the cover on the left to reset. Its intended to be that way so that you know you lost lock Regards Paul WB8TSL On Fri, Feb 6, 2015 at 7:16 AM, Joe D'Elia j...@windrushadv.co.uk wrote: cdelect@... writes: Joe, Nice find. Don't worry about the lamp, they VERY seldom fail. Of course there are electronics failures that crop up. What color is your physics package? Blue paint= old style Olive green paint = mid production Silver (no paint) = late production Also can I get your units serial number for my list? With it in the operate mode watch the control voltage and slowly adjust the fine quartz back and forth. If it is indeed locking to the Rubidium the meter should follow your adjustment. I would not worry about the physics package case either, just put some tape across to seal it from heat loss. I can give you some alignment hints if you need them. Corby Dawson cdelect@... The physics package is silver so you reckon late production. By looking at date codes I reckon it's late 1988 production, does that come into that definition. When did they stop making the unit? Also regarding serial number the tag at the back has been filed so that the number is not legible, however the calibration stickers and a departmental identification tag list it as 916-00184, does that tie in with the rest of the numbers you have? Again from the cal stickers etc this is an ex RAF machine. I switched to control voltage and initially it was off the scale, adjusting the fine control just brought it back to 50, so I tweaked the coarse slightly to take the reading down to 40 and then adjusted the fine as you suggest and the meter follows the direction of adjustment. This also took the output from +25 millihertz to -2 around the 5MHz nominal so it does appear to be locked. The unit does not have any options so just a blank plate on the right and the lock light is red instead of green also the back panel connections are different to any shown in the manuals I've seen, though I have found some of the differences by trawling through the changes section. I'm back to the cal lab next week to pick up a meter that they are calling for me and whilst there I shall see if I can find something of the same vintage that uses the same top and bottom covers and side rails that I can cannibalize to fix this unit. If I can then I'll keep and get it back to pristine condition, if I can't then I'll find someone else who want's to take on the burden. JoeD ___ 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] T.I. questions
On 2/6/15 12:42 PM, Hal Murray wrote: mag...@rubidium.dyndns.org said: The typical noise generator chips uses a PRNG based on DFFs and XOR gate(s). A typical weakness is that the chain of DFFs is to short, causing a relatively high rate of cycling, which hearable as a beating. However, for some uses, that is OK. The buzzword for the typical PRNG is LFSR - Linear Feedback Shift Register. Many years ago, Xilinx published a good app-note on this topic. There is also a section in Art of Electronics. With the right generating polynomial, you get a sequence of bits that doesn't repeat until 2^N-1 bits. The math geeks like to collect them. Each 1 bit in the polynomial turns into an XOR gate, so you will also find collections of polynomials with fewest bits. Only maximal codes have 2^n-1 states/periods. There are other configurations with shorter periods. A particularly tricky thing is that if the shift register in a maximal generator ever winds up as all zeros (e.g. from a upset or bit flip), then the generator sticks at zero. All maximal generators have an even number of taps, too. You can do all kinds of interesting things considering the generator as a polynomial (like factoring) The Spread Spectrum Systems book by Dixon describes lots of ways to make fast generators. The usual vanilla design shows all the taps going to a big modulo 2 adder, but it's more efficient to put an xor between each flip flop, and run the feedback to those stages that have taps. In fact, clever design integrates the XOR into the flipflop logic, so you can basically clock the PN generator at the toggle rate of the flip flops. There are also a plethora of schemes using multiple generators running in parallel with the outputs XORed.. Gold and Kasami codes are good examples where the two generators run at the same rate. GPS C/A and P(Y) code is an example where the generators run at different rates. ___ 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] T.I. questions
Hi Attila, On 02/06/2015 08:30 PM, Attila Kinali wrote: On Fri, 06 Feb 2015 19:29:46 +0100 Magnus Danielson mag...@rubidium.se wrote: Since then interpolation of single-shot events have been investigated, and is now down to 200 fs in the best counter I know of (and have). Which counter is that? I'm only aware of an experimental TDC that does 500fs. Namely the SAW filter based TDC by Petr Panek[1,2], which he presented at EFTS 2013. (I don't have that paper at hand and cannot look it up due to broken german internet). The Wavecres SIA-3000 has a single-shot resolution of 200 fs, but the trigger jitter. Later models only went to 300 fs as far as I know. The DTS-2070C has a resolution of 800 fs. Cheers, Magnus Attila Kinali [1] Time-Interval Measurement Based on SAW Filter Excitation, Panek, 2007 [2] Time interval measurement device based on surface acoustic wave filter excitation, providing 1ps precision and stability, Panek, Prochazka, 2007 http://link.aip.org/link/RSINAK/v78/i9/p094701/s1 ___ 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] HP 5065A questions
On 2/6/2015 4:16 AM, Joe D'Elia wrote: I shall see if I can find something of the same vintage that uses the same top and bottom covers and side rails that I can cannibalize to fix this unit. Joe, You may find the side panels / handles castings since they are common to a lot of that vintage -hp- gear, however the top cover is more or less completely full of ventilation holes which I have never seen in any other instrument of that series, thus perhaps unique to the 5065A. A lot of drilling if you use one from another unit! If you have the 10811 oven then it is pretty certain that it is one of the last few versions of the 5065A so the 916 prefix doesn't sound right. Mine with the 10811 is 2432A prefix for example. A good find, even if you do have to drill a lot of holes! Dan ___ 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] Good references on holdover?
The state of the art 20 years ago is described here: http://www.hpl.hp.com/hpjournal/96dec/dec96a9.pdf They understood their OCXO very very well. And you can find EFC trends on the web for hundreds of different Z3801A's (and similar) if you want to see how the EFC trends (and occasionally jumps). It is very intuitive to watch the smoothed EFC graphs build over time. They were specifically meeting a telco spec for 24 hour holdover. Your 3 second requirement is, um, a bit different! I have no idea what your local clock is, or what you are locking to, if 3 seconds is a long time. Tim N3QE On Fri, Feb 6, 2015 at 4:21 AM, Javier Serrano javier.serrano.par...@gmail.com wrote: Dear all, We would like to start working on holdover performance for White Rabbit [1]. This is a new domain for us. Our main use case is a WR switch losing its reference because someone disconnects a fiber. We can have redundancy, but it will take some time for a switch to change over to another reference. During this time, the oscillator in that switch will be free-running. We want to minimize the phase drift during that interval, which we think should be a couple of seconds maximum. We have never worked on holdover, and I am wondering if we can do something smarter than the obvious feeding of some constant voltage to the VCXO, based on averaging during the locked state. Does anybody know of any good references on holdover? Thanks! Javier [1] http://www.ohwr.org/projects/white-rabbit/wiki ___ 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] Systron Donner IMU/GPS and/or Jupiter Pico
I have a colleague who's using the Systron-Donner MMQ IMU/GPS unit, and he's wondering if there's a way to get integer seconds out of it. It uses a Jupiter Pico GPS, I believe, and one of the messages provides Seconds of Week GPS time, as well as UTC seconds and UTC day, month, year. So here's the question.. he needs seconds from epoch (like TAI time), and it's moderately straightforward to convert D/M/Y to get midnight time from epoch and then add the seconds. The question comes up of whether there's a clever fast algorithm (in C) to get Week time so he can add GPS seconds of week to midnight of week start. Or alternately, does anyone know what the leap second behavior of this unit is, and he can just use the UTC. And in fact, as I sit here before my morning coffee, this brings up an interesting leap second question: if you did a calculate midnight time in seconds and add seconds of day (or week), what do you do for that calculation.. Is the midnight time for the midnight after the leap 86401 greater than the midnight time before leap? That makes calculating seconds of midnight a bit tricky, because it's not just year epoch + month epoch + day of month*86400 ___ 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] Good references on holdover?
On Fri, 6 Feb 2015 10:21:08 +0100 Javier Serrano javier.serrano.par...@gmail.com wrote: We would like to start working on holdover performance for White Rabbit [1]. This is a new domain for us. Our main use case is a WR switch losing its reference because someone disconnects a fiber. We can have redundancy, but it will take some time for a switch to change over to another reference. During this time, the oscillator in that switch will be free-running. We want to minimize the phase drift during that interval, which we think should be a couple of seconds maximum. We have never worked on holdover, and I am wondering if we can do something smarter than the obvious feeding of some constant voltage to the VCXO, based on averaging during the locked state. Does anybody know of any good references on holdover? I think you are looking for something like [1]. I think [2] could be also of help, although it's not as good as the Nicholls paper. Zhou's paper [3] seems to be very similar to what Nicholls did (i have not fully read it yet). HTH Attila Kinali [1] Adaptive OXCO Drift Correction Algorithm, by Nicholls and Carlton, 2004 http://dx.doi.org/10.1109/FREQ.2004.1418510 [2] A Frequency Model for OCXO for Holdover Mode of DP-PLL, by Hwang, Shin, Han, Kim, 2000 http://dx.doi.org/10.1109/SICE.2000.889649 [3] Adaptive Correction Method for an OCXO and Investagion of Analytical Cummulative Time Eror Upperbound, by Zhou, Kunz, Schwartz, 2011 http://www.sce.carleton.ca/faculty/schwartz/abstracts/HuiPaperschwartz.pdf -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ 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] Good references on holdover?
Javier, If you are aim to do hold-over as you switch between two sources, you are looking at reasonably short times, then just keep a fixed voltage to the oscillator suffice. Even if you need a little longer times, say 10-20 s, it suffice. Temperature changes and oscillator drift may be the main problems for longer times. Cheers, Magnus On 02/06/2015 10:21 AM, Javier Serrano wrote: Dear all, We would like to start working on holdover performance for White Rabbit [1]. This is a new domain for us. Our main use case is a WR switch losing its reference because someone disconnects a fiber. We can have redundancy, but it will take some time for a switch to change over to another reference. During this time, the oscillator in that switch will be free-running. We want to minimize the phase drift during that interval, which we think should be a couple of seconds maximum. We have never worked on holdover, and I am wondering if we can do something smarter than the obvious feeding of some constant voltage to the VCXO, based on averaging during the locked state. Does anybody know of any good references on holdover? Thanks! Javier [1] http://www.ohwr.org/projects/white-rabbit/wiki ___ 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] T.I. questions
Bill, A technique similar to this is used in the HP5328A counter, when equipped with the option 040, and when doing the TI averaging. Noise is intentionally added into the 100 MHz control loop, and then multiple measurements is averaged. This way, the sample point moves around it's average, and you can get better average resolution this way. However, the single-shot resolution is still just 10 ns. Since then interpolation of single-shot events have been investigated, and is now down to 200 fs in the best counter I know of (and have). Cheers, Magnus On 02/06/2015 07:52 AM, Bill Hawkins wrote: Referring to a 1952 manual on servo systems, jitter seems to be noise in the system, while dither is intentionally introduced to get a servo through its dead space (usually caused by static friction). The dead space in a counter is the interval between least significant integers. Thus the amplitude of the dither is the size of that dead space at a frequency that will be lost in the average. I can see how this would work for a servo system, but how can a counter display more resolution than its least significant digit? Bill Hawkins -Original Message- From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of Magnus Danielson Sent: Wednesday, February 04, 2015 5:41 PM To: time-nuts@febo.com Cc: mag...@rubidium.se Subject: Re: [time-nuts] T.I. questions Joe, The single-shot resolution of the Option 040 series input is 10 ns. The T.I. interval option adds jitter to make the averaging useful. If your PPS was instead a burst or a clock, comparing it to your clock would make more sense and you could use that higher resolution. If I had a GPIB interface for my 5328A, I would play around with it more. I think I recall a 10 ps resolution being achievable according to fancy specs, with averaging to it's max. Cheers, Magnus On 02/04/2015 09:58 PM, Joseph Gray wrote: Sorry for the simple questions. Old hands can simply ignore this message if it bothers you. I have an HP 5328A that I have pieced together from broken units. As soon as I make a ribbon cable for the GPIB card, I want to use it for T.I. measurements. I know it isn't nearly as good as a 5370, but it is what I have at the moment. The manual states that the Universal module adds dithering to the multiplied 100 MHz internal reference. If I use a GPSDO as an external reference, will this multiplying/dithering allow me to use the same GPSDO (divided down to PPS) on either the START or STOP inputs? If not, then how about using a different GPSDO as the external reference? Or even an FE5680A rubidium? If I read the specs correctly, using T.I. averaging on the 5328A, a 1000 s average should get me 100 ps resolution? With this particular instrument, what is the best, useful resolution I can expect? Joe Gray W5JG ___ 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] question Alan deviation measured with Timelab and counters
Moin Magnus, On Wed, 21 Jan 2015 07:07:54 +0100 Magnus Danielson mag...@rubidium.dyndns.org wrote: For oscillators, they should have been turned on long enough such that any drift is negligible. Alternatively you process out the quadratic trend out of it. The later should be accompanied by some quality measure of how much remaining systematics there is (see Jim Barnes PTTI paper on Drift Estimators). Do you mean The measurement of linear frequency drift in oscillators, http://tf.nist.gov/timefreq/general/tn1337/Tn264.pdf ? Attila Kinali -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ 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] Good references on holdover?
On 2/6/15 1:21 AM, Javier Serrano wrote: Dear all, We would like to start working on holdover performance for White Rabbit [1]. This is a new domain for us. Our main use case is a WR switch losing its reference because someone disconnects a fiber. We can have redundancy, but it will take some time for a switch to change over to another reference. During this time, the oscillator in that switch will be free-running. We want to minimize the phase drift during that interval, which we think should be a couple of seconds maximum. We have never worked on holdover, and I am wondering if we can do something smarter than the obvious feeding of some constant voltage to the VCXO, based on averaging during the locked state. Does anybody know of any good references on holdover? The only thing that might be better is if you have a control loop that forms a model and generates corrections based on that, so that while you holdover, it's not just a constant tuning voltage, but perhaps a slow ramp. For instance, if your control loop has estimated the aging rate (and you have an oscillator that ages really, really fast), you could do that. or if your control loop ingests the temperature and has calculated the temperature/frequency transfer function. But it seems that you have a fairly short holdover time requirement (seconds, maybe a few minutes?) I'm not sure the environment changes all that much in that time span. A simple hold the control voltage might be as good as it gets. I've looked into more complex schemes for situations where you are tracking out something that changes fairly rapidly (Doppler shift for an orbiter) and in a way that isn't a nice linear slope (which could be done nicely by a PID kind of loop). ___ 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] D term (was no subject)
On Mon, 26 Jan 2015 20:15:12 + Poul-Henning Kamp p...@phk.freebsd.dk wrote: The basic math of PID has been around for about 100 years. The invention of the servo (and synchro/resolver) is what makes its day... If anyone wants to dive into control theory I recommend reading the book Feedback control of dynamic systems by Franklin, Powell and Emami-Naeini. It's very accessible, requires only little more than high school math and is geared to an engineering approach (ie lots of real world examples and how to solve them) Attila Kinali -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ 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] Good references on holdover?
Thanks for your ideas. On Fri, Feb 6, 2015 at 11:09 AM, Hal Murray hmur...@megapathdsl.net wrote: Is your PLL analog or digital? I'll assume digital since it's hard to hold analog voltages stable for several seconds. Yes, it is digital. It's even software. It runs on an LM32 [1] soft core inside an FPGA. Then there's a DAC and a VCXO. Our goal is to achieve sub-ns synchronization with respect to a master reference even during switch-over between redundant paths. Cheers, Javier [1] http://www.ohwr.org/projects/lm32 ___ 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] T.I. questions
Pulse width modulation. Suppose the readings go like this: 6,5,5,5,5,5,5,5,5,5,6,5,5,5,5,5,5,5,5,6,5,5,5,5,5,5,5,5,5,5 you would be able to interpolate that result to be 5.1 If it went: 6,6,5,5,5,5,5,5,5,5,6,6,5,5,5,5,5,5,5,5,6,6,5,5,5,5,5,5,5,5 You would be able to interpolate that result to be 5.2 Or: 6,6,6,6,6,5,5,5,5,5,6,6,6,6,6,5,5,5,5,5,6,6,6,6,6,5,5,5,5,5 would be 5.5 and on down the road until you achieved: 6,6,6,6,6,6,6,6,6,5,6,6,6,6,6,6,6,6,6,5,6,6,6,6,6,6,6,6,6,5,... which would be 5.9 I am certain that I have miscounted somewhere (everywhere?), and the digits displayed are not required to be so ridged in how they distribute, just that you are getting ratios between one and the other. -Chuck Harris Bill Hawkins wrote: Referring to a 1952 manual on servo systems, jitter seems to be noise in the system, while dither is intentionally introduced to get a servo through its dead space (usually caused by static friction). The dead space in a counter is the interval between least significant integers. Thus the amplitude of the dither is the size of that dead space at a frequency that will be lost in the average. I can see how this would work for a servo system, but how can a counter display more resolution than its least significant digit? Bill Hawkins ___ 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] T.I. questions
On Fri, 06 Feb 2015 19:29:46 +0100 Magnus Danielson mag...@rubidium.se wrote: Since then interpolation of single-shot events have been investigated, and is now down to 200 fs in the best counter I know of (and have). Which counter is that? I'm only aware of an experimental TDC that does 500fs. Namely the SAW filter based TDC by Petr Panek[1,2], which he presented at EFTS 2013. (I don't have that paper at hand and cannot look it up due to broken german internet). Attila Kinali [1] Time-Interval Measurement Based on SAW Filter Excitation, Panek, 2007 [2] Time interval measurement device based on surface acoustic wave filter excitation, providing 1ps precision and stability, Panek, Prochazka, 2007 http://link.aip.org/link/RSINAK/v78/i9/p094701/s1 -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ 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] Good references on holdover?
Dear all, We would like to start working on holdover performance for White Rabbit [1]. This is a new domain for us. Our main use case is a WR switch losing its reference because someone disconnects a fiber. We can have redundancy, but it will take some time for a switch to change over to another reference. During this time, the oscillator in that switch will be free-running. We want to minimize the phase drift during that interval, which we think should be a couple of seconds maximum. We have never worked on holdover, and I am wondering if we can do something smarter than the obvious feeding of some constant voltage to the VCXO, based on averaging during the locked state. Does anybody know of any good references on holdover? Thanks! Javier [1] http://www.ohwr.org/projects/white-rabbit/wiki ___ 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] Good references on holdover?
javier.serrano.par...@gmail.com said: We have never worked on holdover, and I am wondering if we can do something smarter than the obvious feeding of some constant voltage to the VCXO, based on averaging during the locked state. Does anybody know of any good references on holdover? I doubt if you will come up with anything better. Maybe if you have big temperature changes, it might help to track the recent slope. Is your PLL analog or digital? I'll assume digital since it's hard to hold analog voltages stable for several seconds. I suggest capturing some data from a live system. Say 1 minute or 1/2 hour, whatever you can get with minimal work. Then scan it by eye to get a feel for what it does in normal operations. Then maybe patch your collector to only collect data for big values of the types of stuff that look interesting. Do you see any jumps? What's the biggest change over your worst case holdover? (or a bit longer for caution) I'm assuming you can compute or measure the frequency change per count. Do you see anything that will go over your limits? I assume you have some target limits rather than are doing this just because it feels good. If not, compute the predicted limits and see if your users are happy with that. One thing to watch out for is recovery. Do you have something like a PPS? If so, you have a decision to make. Do you want the PPS back to the correct timing as fast as possible or can it get there slowly? The PPS fixup time will be the integral under the frequency offset curve. You can have a tall narrow spike, or a short wide bump. For the spike, the frequency will be way off, but not for long. For the bump, the frequency will always be close, but it will be off for a long time. -- These are my opinions. I hate spam. ___ 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.
