> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of
> donald collie
> Sent: Wednesday, April 18, 2018 2:40 AM
> To: Discussion of precise time and frequency measurement
> Subject: Re: [time-nuts] 4046 replacement
>
> Thankyou Attila. I remember reading a
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of John
> Ackermann N8UR
>
> Reviving the conversation about superb voltage regulators, I am looking
> for one to run the analog and PLL bits of a high performance frequency
> synthesizer chip.
>
> The current drain looks to be about
Request for data from an off-list friend who is looking for info on an Ovenaire
OCXO and/or the instrument it was used in:
> Could you send a post on my behalf asking if anyone else has an
> Ovenaire 42-15 or if not, a Spectracom 8131 Frequency Standard Oscillator.
This is from Dennis Tillman,
> To echo Attila, just sayin.
$4000 is up there, but not outrageous IMO. If you have a $4000 budget and
don't actually need a primary standard, a well-kept 5065A is a good way to go.
At taus less than a few hours they can outperform the best 5071As, even before
Corby's mod is applied. An
> Hi
>
> Well, if anybody else is “giving away” 5065’s I’d certainly be willing to
> “accept”
> one :)
>
> Bob
Be careful what you wish for. :) The unit Pete is talking about is a 1969-era
model, in nice overall shape but with the usual bad capacitors on the oven
controller board. The lamp ov
> -Original Message-
> There is a small detail that puzzles me: the ADEV for the internal
> reference 10MHz and 30MHz bends upwards at tau greater than 0.1s.
> Shouldn't that be visible by a change in slope in the phase noise
> plot at ~10Hz as well?
>
> Attila Kinal
rop me a note offline and I'll hook you up.
-- john, KE5FX
Miles Design LLC
> What are the specs ? 73 de N1UL
>
> Sent from my iPhone
>
> > On Jan 27, 2018, at 5:38 PM, John Miles wrote:
> >
> > So, who's the lucky winner? Anyone on here? That's
So, who's the lucky winner? Anyone on here? That's the first one of these
I've seen in the "secondary market," so to speak.
https://www.ebay.com/itm/332531180078
-- john
Miles Design LLC
___
time-nuts mailing list
> Yes, exactly those. With no input, would they be expected to be 4 kHz off?
> The spec for the standard part wants the input to be within 1e-7. I would
> not expect them to be so far off free running.
That does seem like a lot. I'd expect a few hundred Hz of error at the most.
> I saw 13 MHz o
Sounds like he's talking about the small 'bricks' that Wenzel sells with
internal PLL-disciplined OCXOs. Some of these expect oddball input
frequencies. Just looking at the 80 MHz parts on the shelf around here,
500-14273 wants a 13 MHz input, 500-25010 uses 24.576 MHz, and 500-25009 uses
19
> But be aware, that measurements close to the limit of thermal noise
> will make your measurement go sour. There the noise of your splitter
> will cause an anti-correlation effect and the measured noise will
> suddenly drop way below thermal noise. Craig Nelson and Archita Hati
> from NIST, Enrico
I don't believe there's a way to save images from within HTBasic, although
there certainly should be. I use the "Snipping Tool" accessory in Windows 7
for that sort of thing. It probably exists in all versions of WIndows
including 10, but who knows what they call it these days.
If you crea
> One of the few things I dont have a manual for
>
> Can anyone tell me if there is an HP model number for the Reference
> Oscillator Assembly (A19) in the 5071A? Might be a a 10544A ?
>
The OCXO was 05071-60219. As Rick says, it was a modified 10811. The 10544
was long gone by the 5071A's
> Do you recall what the amplitude of the input signal to the Wenzel shaper
> was?
>
> Since I used a 1:2 (3dB) terminated splitter for my measurements the input
> to the splitter is the same as the Timepod reference signal amplitude
>
I was using a similar setup, so the shaper input would have
> I have measured the PN of the LTC6957-4 at 10MHz.
>
> I could measure the PN of the TAPR variant of the Wenzel circuit as well as
> the PN of the comparator based circuit (with CMOS output buffer).
>
These plots came from the Wenzel diff-amp shaper:
http://www.ke5fx.com/wenzel_resid_PN.png
htt
> Thanks for clarifying this issue! I may use references with close-in PN
> (1Hz-1kHz) 10 dB noisier than the DUT, however as you said it may
> require an overnight run for 1 Hz offset, which isn't the nice part... I
> assume then that speeding up the measurement process can only be
> obtained usin
> In a PN measurement system with a dual channel cross correlation is there a
> simple rule of thumb for how low should be the PN of the (external)
> references compared to the DUT? (even with a 20 dB noise floor
> improvement with 1 cross correlations)
If you have two references, then it's OK
> My intention is to leave the instrument on for severalweeks in order to see
> if it is stable. Right nowmy concern is wether the 2nd harmonic
> readingindicates a wear out of the photo cell.
> When repairing the RFV Cavity, all components weretaken out and carefully
> examined. There wereno dimme
I'd recommend a GPIB-USB-HS adapter. There are tons of them on eBay, e.g.
http://www.ebay.com/itm/322360711342
http://www.ebay.com/itm/321957086926
These will work well with both PN3048 and HTBasic. I've purchased them from
both of those sellers in the past. They appear to be genuine, unlik
13, 2017 8:41 AM
To: Discussion of precise time and frequency measurement; John Miles
Subject: Re: [time-nuts] Three-cornered hat on timelab?
Hi John,
Thanks! With lesser equipment, such as the 5370A, would longer runs be better?
I used a set of 1 hr runs and the result wasn't
> It was mentioned that Timelab can do a three-cornered hat. I can't find it.
> Is
> this something that can only be done with multiple Timepods connected, or
> is there an option that I'm missing? Or is there some other tool that needs
> to be used? For the record, I want to create the 3c-hat
> We are having trouble getting beam current on our #2 HP5061B. Ion
> current has gone down in a week from 10 to zero. It looks like the rf
> chain is working normally. Both HV power supplies had to repaired.
> Beam current is all the way up and which gives 2 on the meter. We are
> preparing to d
Try running the 32-bit version (timelab.exe) instead of timelab64.exe, even if
the VM supports 64-bit execution. That can sometimes help with compatibility.
-- john, KE5FX
Miles Design LLC
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of Bob
> Bown
Yeah, I wouldn't worry about the oven, but the ionizer filament is a different
story. Those can definitely open up.
-- john, KE5FX
Miles Design LLC
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of
> Donald E. Pauly
> Sent: Wednesday, March 22, 20
That's some very nice work, Donald. Looking back, I have junked one or two Cs
tubes that might have been usable if I'd thought through the problem of high
ion pump current as you and KB7APQ seem to have done.
Another good reason to raise the lockout threshold would be to cut down on the
repet
I'll add MB_DEFBUTTON2 to those prompts to keep that from
happening.
If Wine is interpreting a second Escape keypress as 'Yes', then that's
definitely a bug on their part. Worth reporting to them if it still happens in
the current version.
-- john
Miles Design LLC
__
es, it does exit.
> Bob
Hmm. I can't repro this on Win7 x64 -- what version of Windows are you running?
If I hit Escape followed by Enter, it does exit, since 'Yes' is the default
choice. But hitting Escape twice doesn't do anything (and shouldn't).
-- john
Miles Des
Agreed, it's probably reasonable to say that a real absorption wavemeter would
(a) have to have a meter or some other visual indicator; and (b) likely be
powered exclusively by the energy its tank circuit "absorbs." The BC221/LM
boxes fall a little short of both requirements.
-- john, KE5FX
You could use it as an absorption wavemeter, in its broadest sense of a passive
tuned circuit with an indicating load. The headphone jack was normally used to
calibrate the VFO against a harmonic of the internal crystal oscillator, but it
could zero beat an external source as well. It stretche
> Search the list for what Paul and I did. There is a web page where someone
> did the same forced pump down. If all the voltage s are good you have
> nothing to loose. You are going to need a "real" high voltage source.
That may have been my page here: http://www.ke5fx.com/cs.htm
Bertan HV supp
> What is the general feeling here about this issue? I confess that if the amp
> output is transformer coupled, I see exactly zero benefit in a grounded
> connector as the feed from the amplifier.
This question comes up every so often. It comes down to whether you want your
test setup to look l
You may find that the behavior varies quite a bit depending on what you put
between the maser output and the doubler. I've seen one case in particular
that generates a ton of PM spurs, specifically an HP 11721A doubler driven by
the 5 MHz output of a 5061A Cs standard. The 5 MHz output uses a
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of John
> Ponsonby
> ... It turns out that the resonant frequency of the cavity is much more
> critically dependent on its diameter than on its length. So it would be best
> to be able to mount the bulb in
> Inspired by Corbys post, I thought I'd follow up with putting the
> telemetry-data from EFOS 3 on the web, for those with any interest in these
> things. There is also a small picture gallery. Can be found on
> www.efos3.com.
> If I got the scheduling right, weekly plots should update every 5 min
Sorry about that. :) Send me a new zipfile and I'll update the distribution...
-- john, KE5FX
Miles Design LLC
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of Mark
> Sims
> Sent: Monday, December 26, 2016 10:12 AM
> To: time-nuts@febo.com
> Subjec
The error message in question came from the IPCONN::connect() method in
ipconn.cpp. That particular code (11004) is returned by gethostbyname() when
it fails to resolve a DNS name. Normally, when Heather tries to connect to the
test server, gethostbyname() is used to turn "ke5fx.dyndns.org" in
Hmm, no, that shouldn't affect anything. Sounds like a DNS issue of some kind,
presumably (but not necessarily) at your end. Try again?
-- john, KE5FX
Miles Design LLC
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of
> Richard W. Solomon
> Sent:
> As said, if I put the .cfg file in the system32 directory I can set eg.
> the new time zone.
> Strange, if I enable eg. the esc esc command to shut off LH, it does the
> job well, but with the exception for the time zone setting, this and all
> the other commands once set are not anymore to be di
Yeah, that's probably something we should fix. Installation programs don't
normally set that field at all. But when Heather's initial working directory
isn't in the usual Program Files hierarchy, it tries to load the .cfg file from
the same directory as the executable. This is something I do
Hmm, that doesn't make any sense at all. Nothing should be written to
\windows\system32 by the installer under any conditions.
-- john, KE5FX
Miles Design LLC
> Many thanks Brian!
>
> I found it!
> In Windows 7 (10?) this file must be copied into the system32 directory
> of windows in order
> It seems during the install at least on Windows 10 that it does not give the
> full permissions to the files. I just gave the folder the correct permissions
> after the isntall. If you go to help, at the bottom of the screen it tells
> you
> where during the install another copy of the heather.
> thanks for the offer, Geo
> We've already found a new Prologix unit at much less than retail. I'm not
> familiar with the NI units, and have seen postings that they are
> tempermental with some instruments.
I've mentioned before that NI adapters are the _least_ temperamental of all of
them.
If I remember correctly, the oven power supply runs off of the same bus that
powers the rest of the unit, in the 5062C's case. It took a fair bit of
hacking to install the switch and make it do what I wanted. Not much besides
the ion pump and the OCXO remains active with the switch in the 'of
> 1. Cesium depletion, which only occurs when the tube is operating with
> cesium oven on, and high voltage at the other end; and
The Cs atoms need to be electrically neutral, so their depletion rate shouldn't
depend on the presence of HV, just the oven heater. State selection wouldn't
work on
Yep. It should be all good now (once the DNS records propagate.)
-- john, KE5FX
> Domain name expired on September 8th - needs renewing!
>
> On 11/09/2016, at 10:48 AM, Nick Sayer via time-nuts wrote:
>
> > I tried to look for Lady Heather docs today, but it appears like the ke5fx
> domain is…
Hmm. Gotta be some kind of DNS glitch, it's not due for renewal as far as I
can see. I'll look into it, thanks.
-- john, KE5FX
Miles Design LLC
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of Nick Sayer
> via time-nuts
> Sent: Saturday, September
> That's a pretty scary problem to have. Are these frequency counts or TI
> readings? You wouldn't normally see two identical TI readings in a row,
Actually, that's not even safe to assume for TI readings, depending on how your
triggering works. It would make sense to do whatever it takes to f
> I've got a file with counter values that are latched once per second,
> with the count read from the latch every half second. So, generally,
> there are two identical values, then two different identical values,
> etc. But, of course, the routine that does the every half second
> reading isn't
> I am just starting out using Lady Heather.
>
> I am trying to create a .cfg file to set the program to COM 3 at startup.
>
> I created the file and saw the instructions to put it in my “Documents”
> folder.
> The only documents folder in Windows is “My Documents.”
>
> I created a text file an
> If the oscillator output is thru the resonator, then at large offsets,
> the source impedance is reactive. It can easily have an effective
> temperature less than room temperature. If this "source" is then
> used with a low noise temperature preamp, it is entirely possible
> to go beyond these
-177 = the -174 dBm/Hz SSB thermal noise floor at 25C, less 3 dB to account for
the usual assumption that half of it is AM, half PM.
dBm/Hz is obviously equivalent to dBc/Hz for a 0 dBm carrier.
Anyone who claims to measure noise in a 1 Hz bandwidth below -177 dBm/Hz at
room temperature is
Good point, the 'milliwatt' part of 'dBm' takes the E^2/R part out of the math.
If we were speaking of dBv/Hz, the system Zo would need to be considered to
determine the power.
-- john, KE5FX
Miles Design LLC
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com]
Right, I'm speaking specifically of L(f). The device being driven by the
oscillator doesn't care about the NF of the driver stage, only what a PN
analyzer would measure at the output jack.
For any 50-ohm source, the practical L(f) floor is -177 dBm/Hz - the carrier
power in dBm. No oscillator
Or rather -(177+DUT output power in dBm). The minus sign makes the difference
between the thermal floor and a nuclear war!
-- john, KE5FX
Miles Design LLC
> Remember that L(f) is expressed in dBc/Hz, not dBm/Hz. If it were dBm/Hz,
> then kT would be the limit. But in dBc/Hz terms, the limit i
Remember that L(f) is expressed in dBc/Hz, not dBm/Hz. If it were dBm/Hz, then
kT would be the limit. But in dBc/Hz terms, the limit is 177 + the DUT's
output power in dBm.
Assuming a 50 ohm system, of course.
-- john, KE5FX
Miles Design LLC
> -Original Message-
> From: time-nuts
>
> Very selected and incomplete references and the equally important
> question
> of measurements strangely not covered
>
> 73 de N 1 UL
>
I suppose he could write an equally-lengthy article on measurements alone, but
leaving out the post-1970s history entirely was a little disappointing.
> I've had a specific GPSDO running for some time now, and I notice that the
> noise at tau 1s has gotten worse as the retrace flattens out. In this case,
> the ADEV was about 3.6E-11 a month or so ago, and has now gone up to about
> 8.5E-11. (Measurements performed by a 5370A against a PRS-45A
> But TimeLab doesn´t list any GPIB interfaces.
>
> Did anyone have success with this combination? Any insights
> highly appreciated...
Agilent's NI488.2 compatibility layer seems to work only with 32-bit apps, not
64-bit ones, so you may be able to get it running by modifying your Windows
shor
> On 6/18/16 6:17 PM, Tom Knox wrote:
> > Does anyone have any insight as to why Microsemi has stopped shipping
> 3120A Phase Noise test sets beside that they are not to spec?
> >
> They're shipping them.. I got quoted a 12 week delivery in April, $9k
> for the box, about $12k-15k for the software
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of Tom Knox
> Sent: Saturday, June 18, 2016 6:17 PM
> To: Time-Nuts
> Subject: [time-nuts] 3120A issues
>
> Does anyone have any insight as to why Microsemi has stopped shipping 3120A
> Phase Noise test se
> I note your software works with the HP 7 too. As I say, I already have
> one of them, but was thinking the E4406A would be a worthwhile addition.
> But perhaps not in that case. Although maybe it will outperform my 7
> system for phase noise measurements. I expect you will know the answer
> and looked at what was used to make the phase noise measurements. It was an
> Agilent E4406A. The noise floor is nowhere near as low as the more
> expensive instruments, but the E4406A is available for under $500, which is
> more than two orders of magnitudes cheaper than an E5052B.
It appears t
> I was thinking more of the sidelobes: if you're looking at a quiet
> oscillator (e.g. -140dBc @ 100Hz) , with a 1 second epoch, and you want
> to measure the noise at, say, 100Hz out, the window function needs to be
> down 140 dB at that bin.
>
> WIndows like uniform and Hamming are probably onl
> What sort of windows do folks use for making FFT measurements of phase
> noise.
>
> Say you have 1 second of sampled data (so the FFT resolution is 1 Hz).
> If you're interested in the noise power at, say, 10 Hz away, a
> rectangular window isn't going to be very far down, unless you have a
> LO
One workaround for the 1-million point limitation on imported data is to use
"Acquire->Acquire from live ASCII file" instead of "File->Import ASCII
phase/frequency data." Most of the same code is used for both cases, but
unlike the static file-import version of the dialog, the live data importe
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of Mark
> Kahrs
>
> Anyone else going to IMS in SFO this coming week?
>
>
> I would propose meeting during the "industry Reception" Wednesday
> afternoon...
I'm there from Monday night through Thursday, and that sounds like a good
Public service announcement: I just tried to stream this episode on Amazon
Prime Video, and it seems they have some sort of indexing snafu going on. They
offer episode 1 ("Can We Time-Travel?") and episode 2 ("Are We Alone?") for
viewing, but what actually played when I selected them were two c
I've updated my existing LH download page with Mark's new beta version in both
.zip and (Windows) setup.exe formats:
http://www.ke5fx.com/heather/readme.htm
There's also a link to his readme.txt file with Linux build instructions.
(Mark is the best point of contact for anyone with questions on
> My signal generator has a poor, intermittent power supply. The RF
> section is ok. Who can please tell me which company can and will fix these
> older
> but excellent generators?
>
> Thanks, Ulrich
Have a look at the capacitors in the voltage divider that drives the bases of
the switching
> Here it takes slightly less 3 minutes, but stops at 10M samples.
> Memory consumption of Timelab stays below 350MB the whole time
> and drops to 98MB after it finished.
>
> OS is windows XP pro
>
> How many samples did you get? It should have been 14926518 (ie slightly
> less than 15M)
> > BTW: I discovered that Timelab stops processing after 10'000'000 datapoints,
> > which is kind inconvenient when doing a long term measurment...
>
> I didn't know that. Good to know.
Attila, wasn't this related to an invalid ':' character in the filename coming
through from VirtualBox? Or i
> BTW: I discovered that Timelab stops processing after 10'000'000 datapoints,
> which is kind inconvenient when doing a long term measurment...
It had better not! :) Any steps to reproduce?
-- john, KE5FX
Miles Design LLC
___
time-nuts mailing list -
> I'm looking for off the shelf connectorized amplifiers with very good
> reverse isolation for doing things like intermod measurements (e.g. 2
> signal generators followed by amps/pads into a combiner) and phase noise
> measurements for a digital receiver.
>
> 1-100 MHz kind of frequency range..
My understanding from someone who has purchased one recently is that the
frequency offset error was indeed just a bug, and that it's now fixed. I've
added a note to the end of my page at http://www.ke5fx.com/gpscomp.htm to that
effect.
If the bug has been fixed (or the algorithm replaced), and
Hi, Timothy --
> I've had an Agilent 53220A for some time, and recently discovered the
> wonderful TimeLab software, but I've hit a bit of a snag trying to run some
> ADEV measurements on a Rb clock. The problem is that the graph doesn't
> line up with the amount of data TimeLab says is collecte
> Therefore, talk-only mode is a big advantage in terms of decoupling
> on RS-232 and makes almost no difference on GPIB.
That's not the case when it comes to counters. By timing issues, I wasn't
referring to layer-1 handshaking, but rather the interplay between the GPIB
software application, t
Agreed with Magnus that there are a lot of possible variables in your setup
that need to be ruled out.
Are you using the SR620 driver in TimeLab, or did you find a way to get it to
emit data continuously via the RS232 port for use with the talk-only driver?
I've seen occasional instances whe
16 8:36 am, Chuck Harris wrote:
> > Unfortunately, LH uses a graphics toolkit that was written by
> > John Miles, and it, and he, is windows only.
>
> Even compiling as just the server only still compiles and links the
> graphics toolkit?
>
> --
> Chris Caudle
>
>
> Amp (now Tyco) did introduce them. Called "Decoupled Connectors'' Info at
> http://www.te.com/usa-
> en/search.html?q=Decoupled%2BConnectors&source=header
>
> Tyco parts 413476-2 and 5413476-2 are available from Mouser, maybe Digikey
> as well
>
And you might notice that the TimePod itself d
You don't necessarily need a lot of shielding, at least not at these
frequencies. The decisive battle against EMI is fought at the board level. If
you lose it there, then all you can do is use various patches and hacks,
including external shielding, to cover your retreat.
Given what looks l
Those spurs are reminiscent of what happens when you lift the ground of a coax
cable at one end and turn it into an antenna, in my experience. It is almost
always a bad idea to do this. Try shorting out the capacitor(s) at your input
and output jacks.
-- john, KE5FX
Miles Design LLC
> My '
> It seems there is no way to estimate frequency from looking at the phase
> data then?
Not to my knowledge. The frequency count chart on the instrument itself is
always correct, since it can take the internal baseband offset out of the
incoming data. Because TimeLab doesn't have access to
> Microphonics.
>
> What's the norm here? I see it in the 2nd harmonic level on the panel
> meter. It runs about 20 normally, but a pretty light tap (with my
> fingernail) anywhere on the instrument can peg the meter for a few
> seconds. While I had the scope on A8 TP2 / TP3 (showing the off-re
> Hi all,
>
> A 5115A phase noise test set landed in our lab and I am wondering about the
> data collection.
> It has two telnet ports one for commands and one for phase data.
> When issuing "start" on the command-port it starts spitting out
> phase-difference values on the data port.
>
> However
The reference was probably a bit too obscure for an international audience. :)
https://www.youtube.com/watch?v=VPJqIT7a3qA
-- john, KE5FX
Miles Design LLC
> Hi John,
>
> Thank you very much for this explanation, I found it very "explicative".
> What I am not able to grasp is the sense of the
explanation, thanks John!
>
> On November 27, 2015 2:54:51 PM CST, John Miles wrote:
> >So, here's how I finally grokked this stuff. c, the speed of light in
> >a vacuum, is often spoken of as a "speed limit" that nothing can ever
> >exceed. That's a bad
So, here's how I finally grokked this stuff. c, the speed of light in a
vacuum, is often spoken of as a "speed limit" that nothing can ever exceed.
That's a bad way to put it, and people who have expressed it that way in
popular science writing for 100 years should feel bad.
Instead, the wa
> -Original Message-
> From: time-nuts [mailto:time-nuts-boun...@febo.com] On Behalf Of Jason Ball
> Sent: Friday, October 23, 2015 10:38 PM
> To: Discussion of precise time and frequency measurement
> Subject: [time-nuts] Timelab Query (likely noob error)
>
> Hi all,
>
> I'm afraid this
SRS likes to isolate their RF connectors from ground, it seems. This is almost
never a good idea. When measuring PRS10 Rb standards, I've seen similar line
spurs that went away when I grounded one side of the balanced 10 MHz output at
the point where the DC supply was connected to the unit.
If you're only loading two files, that's definitely not going to work since a
3-cornered hat requires 3 plots. In your example, there must be a third plot
representing A-C, so that each of the three sources will contribute to two
different plots.
Apart from that, check for mismatched file pa
> Hello all - I have always wanted an HP 3048A system. I have acquired a
> 11848A, but need an HP 3561A or equivalent.
>
> Is there any substitute to the 3561A?
Nope. It's the 3561A or nothing.
> Also, are their other spectrum analyzers that are acceptable other than the
> ones listed here?
>
One idea would be to leave the existing C-field circuit in place with a switch
to disable it. That'd preserve your ability to carry out the second harmonic
adjustment procedure, while allowing you to switch back to a stable current for
normal operation.
For that matter, it's not clear what "ag
> On that note, did you look more closely on the NIST analysis of
> cross-correlation and a possible cancellation and thus overly optimistic
> results? Did it have any consequence on your code? What did you take
> away from it?
Yes, you can definitely get divots in the PN trace, especially in mult
> The is a publication by the NPL that talks about a similar technique, but
> comes to a somewhat different conclusion.
>
> http://publications.npl.co.uk/npl_web/pdf/mgpg68.pdf
>
> Bottom line: it holds that the phase noise of all three sources can be
> determined if they are taken two at a time;
s. Frequency drift between Ch0 and
Ch2 will also tend to keep the spurs from cancelling, and will also make it
hard for the software to remove them. It's a good idea to use clean, stable
sources for these types of measurements whenever possible.
> Just a personal note. The Timepod is
Actually, after typing all that, it occurred to me that you might have actually
meant to ask about N-cornered stability measurements. They aren't supported by
the current official release of TImeLab; you need to use the beta version from
http://www.miles.io/timelab/beta.htm .
For instruction
One nice thing about phase noise is that it's computable with complex FFTs,
rather than the one-dimensional phase or frequency differences that ADEV uses.
Another nice thing is that it's stationary -- meaning its probability
distribution can (usually) be treated as unchanging from one measureme
I think the simplest way to explain the evils of TI averaging is that white
noise doesn't alias in a conventional sense. If a value is perfectly random,
then it doesn't matter how you sample it. Your sampling bandwidth -- and
nothing else -- determines how much energy you get. You can legitim
> Hi
>
> If that data is correct, then the 10811 you have is defective.
> Bob
Well... some of the data is reasonable for a scenario where a counter is being
used to measure OCXOs.
Looking at the ADEV plot, I'd say the blue trace (HP105B vs 5065A) is the most
questionable one if it came from
> Apart from the delayed measurement point, I have not been able
> to identify any issues.
>
> The frequency spectrum filtered out by the averaging is wy to
> the left of our minimum Tau.
>
> Phase wrap-around inside bursts can be detected and unfolded
> in the processing.
>
> Am I overlook
1 - 100 of 1088 matches
Mail list logo