Re: [time-nuts] OXCO Spurious Output at Line Frequencies

[Bill Hawkins]

Just to clarify - AC motors are either synchronous or induction.
Induction motors must slip away from line frequency to develop the
magnetic field they need to carry a load. The name plate speed is for
rated load and the number of magnetic poles in the rotor. A 2 pole motor
spins near line frequency - 4 poles at half. 1750 is for a 4 pole motor.
And so on ... 

Bill Hawkins 

-Original Message-
From: Hal Murray
Sent: Tuesday, July 12, 2016 7:24 PM

Most of the medium sized motors run at slightly below line rate, and
often at half of line rate.  I'm thinking of the classic squirrel cage
induction motor vs a synchronous motor.  A quick google search found
1725 and 1750 RPM.


___
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] OXCO Spurious Output at Line Frequencies

[Mark Sims]

A friend on mine once worked on  projects to build very sensitive magnetometers 
(submarine detection and space probes).  Their test lab was in the middle of a 
square mile of land selected for its low magnetic residue properties.   The 
building was made completely (of carefully selected) wood... no nails, etc.   
Everything that went into the construction was scanned.

You had to hike into or be ferried into the lab... no cars allowed anywhere 
near the place.   You also had to wear specially selected clothes (no metallic 
fasteners, residue, dyes, etc).   No metal eyeglasses, etc allowed.   No 
wallets (credit cards) or currency allowed (magnetic ink,  the things could 
detect a dollar bill across the room).  No amalgam dental fillings or medical 
implants allowed.  I always wondered how they allowed any test equipment, 
electricity, etc into the place...


Unfortunately I don't have a cornfield or equivalent readily available, 
although there is a carpark and paddock at the local zoo to which I have access 
after hours should I want.
___
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] OXCO Spurious Output at Line Frequencies

[Bruce Griffiths]

I can't seem to get the line related spurs much below -170dBc/Hz for the 
fundamental and -180dBc/Hz for the line frequency related harmonics. Everything 
was bolted down to a 12mm thick aluminium slab. Avoiding an Earth loop by 
isolating the Timepod Ch0 and CH2 inputs from the aluminium plate seemed to 
help.
Unfortunately I don't have a cornfield or equivalent readily available, 
although there is a carpark and paddock at the local zoo to which I have access 
after hours should I want.
Bruce


  From: Bob Camp 
 To: Discussion of precise time and frequency measurement  
 Sent: Wednesday, 13 July 2016 9:53 AM
 Subject: Re: [time-nuts] OXCO Spurious Output at Line Frequencies
   
Hi

The “practical” solution often is to ship it to an office of the same company 
that happens
to have 60 Hz power. The spur moves to 60 / 120 / 180 and they move on with 
their
evaluation. 

I have also played the “let’s do it with batteries" approach. Screen rooms were 
of little benefit.
Simply running test gear on batteries did not do the job. Ultimately we wound 
up in the 
middle of an Illinois corn field with a bunch of gear modified to run purely on 
batteries. The
spur did go down, but it never fully went away.  It’s been 40 years since that 
adventure so I 
really could not say just how far down we got it …. 

Anything past -120 db at line frequency is in the “don’t bother” category. 

Bob


> On Jul 12, 2016, at 6:44 AM, Martyn Smith  wrote:
> 
> Hello,
> 
> I have a customer who is measuring the phase noise of my 10 MHz ultra-low 
> phase noise frequency standard.
> 
> He is seeing spurious signals at line frequencies (50 and 100 Hz as we are in 
> Europe) at a level around -130 dBc.
> 
> My opinion is that it's impossible to get much better than that.  Even 
> running on batteries make little difference, since the equipment is in a test 
> rack with AC signals everywhere.
> 
> Even the £50k R test set he is using only quotes a spurious spec of -90 dBc.
> 
> What experience does anyone have here?
> 
> Best 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.


  
___
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] OXCO Spurious Output at Line Frequencies

[Bob Camp]

Hi

ummm …… errr ….. yes. 

If you mount the OCXO that drives an 800 MHz FM narrowband transmitter in the 
middle of a wall panel on the rack, you *can* measure the rotation speed of the 
blowers.
That’s not a guess, that’s empirical knowledge :)

Bob

> On Jul 12, 2016, at 8:24 PM, Hal Murray  wrote:
> 
> 
> kb...@n1k.org said:
>> All of those motors in the building are rotating at line frequency. You
>> don’t just have an  electromagnetic field. You can have an acoustic /
>> vibration field as well. Probably not a big  deal, but it’s there …. 
> 
> Most of the medium sized motors run at slightly below line rate, and often at 
> half of line rate.  I'm thinking of the classic squirrel cage induction motor 
> vs a synchronous motor.  A quick google search found 1725 and 1750 RPM.
> 
> Is it possible to measure things accurately enough to notice that?
> 
> 
> -- 
> 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.

___
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] GPS for Nixie Clock

[John Swenson]

Hi Hal,
I've done NTP before, I wanted to do this with GPS since I have never 
done a GPS implementation before.


Xerox Alto, wow that brings back memories! My dad worked at Xerox in 
Palo Alto for 20 years in the 70s and 80s. He primarily worked on the D 
machines but I did get to use an Alto a few times when I was in high 
school. I learned Mesa sitting in front of one.


That must have been 3Mbit Ethernet right?

I remember going to the lab one Saturday, they were working on a machine 
that was supposed to have a custom chip, but it was not ready, so they 
used 7 ASICs, but those were not ready yet either, so they made a large 
board out of TTL for each ASIC with a ribbon cable into the motherboard, 
these boards were hung on fishing line from the ceiling so they were 
just above the motherboard with a very short cable. That thing really 
looked like something out of a movie.


John S.


On 7/12/2016 1:21 PM, Hal Murray wrote:


t...@patoka.org said:

In addition, even MCU has not enough resources to handle TCP/IP, DHCP  and
NTP, it is some solutions available to outsource it to dedicated  chips. I
was using WIZ5100 (assembled as a modules) with great success.


NTP is pretty simple.  If you are willing to take a few shortcuts, things get 
even simpler.  If you have a NTP server on your LAN, you can skip routing and 
DHCP by broadcasting a request.  It would be interesting to see how small you 
could make a set-the-time package.

The Xerox Alto had an Ethernet boot loader in microcode.  (and space in the 
microcode ROM was tight)  When bringing up a new machine, we figured out that 
it was much simpler and faster to debug the Ethernet before the disk 
controller.  The debugger worked over the Ethernet.  That gave you a solid 
place to stand early in the debugging process.




___
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] OXCO Spurious Output at Line Frequencies

[Hal Murray]

kb...@n1k.org said:
> All of those motors in the building are rotating at line frequency. You
> don’t just have an  electromagnetic field. You can have an acoustic /
> vibration field as well. Probably not a big  deal, but it’s there …. 

Most of the medium sized motors run at slightly below line rate, and often at 
half of line rate.  I'm thinking of the classic squirrel cage induction motor 
vs a synchronous motor.  A quick google search found 1725 and 1750 RPM.

Is it possible to measure things accurately enough to notice that?


-- 
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] OXCO Spurious Output at Line Frequencies

[Bob Camp]

Hi

Bottom line:

This is why you often see spur specs stated with a lower frequency limit (like 
250Hz) or 
an exclusion for line related spurs. 

One really fun one not (yet) mentioned: 

All of those motors in the building are rotating at line frequency. You don’t 
just have an 
electromagnetic field. You can have an acoustic / vibration field as well. 
Probably not a big 
deal, but it’s there ….

Bob

> On Jul 12, 2016, at 6:45 PM, Alan Melia  wrote:
> 
> Bob the VLF guys in Europe can receive the 60Hz signal over here! so you 
> probably needed a test site on the far side of the Moon :-))
> Alan
> G3NYK
> 
> - Original Message - From: "Bob Camp" 
> To: "Discussion of precise time and frequency measurement" 
> 
> Sent: Tuesday, July 12, 2016 10:53 PM
> Subject: Re: [time-nuts] OXCO Spurious Output at Line Frequencies
> 
> 
>> Hi
>> 
>> The “practical” solution often is to ship it to an office of the same 
>> company that happens
>> to have 60 Hz power. The spur moves to 60 / 120 / 180 and they move on with 
>> their
>> evaluation.
>> 
>> I have also played the “let’s do it with batteries" approach. Screen rooms 
>> were of little benefit.
>> Simply running test gear on batteries did not do the job. Ultimately we 
>> wound up in the
>> middle of an Illinois corn field with a bunch of gear modified to run purely 
>> on batteries. The
>> spur did go down, but it never fully went away.  It’s been 40 years since 
>> that adventure so I
>> really could not say just how far down we got it ….
>> 
>> Anything past -120 db at line frequency is in the “don’t bother” category.
>> 
>> Bob
>> 
>> 
>>> On Jul 12, 2016, at 6:44 AM, Martyn Smith  wrote:
>>> 
>>> Hello,
>>> 
>>> I have a customer who is measuring the phase noise of my 10 MHz ultra-low 
>>> phase noise frequency standard.
>>> 
>>> He is seeing spurious signals at line frequencies (50 and 100 Hz as we are 
>>> in Europe) at a level around -130 dBc.
>>> 
>>> My opinion is that it's impossible to get much better than that.  Even 
>>> running on batteries make little difference, since the equipment is in a 
>>> test rack with AC signals everywhere.
>>> 
>>> Even the £50k R test set he is using only quotes a spurious spec of -90 
>>> dBc.
>>> 
>>> What experience does anyone have here?
>>> 
>>> Best 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. 
> 
> ___
> 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] OXCO Spurious Output at Line Frequencies

[Alan Melia]

Bob the VLF guys in Europe can receive the 60Hz signal over here! so you 
probably needed a test site on the far side of the Moon :-))

Alan
G3NYK

- Original Message - 
From: "Bob Camp" 
To: "Discussion of precise time and frequency measurement" 


Sent: Tuesday, July 12, 2016 10:53 PM
Subject: Re: [time-nuts] OXCO Spurious Output at Line Frequencies



Hi

The “practical” solution often is to ship it to an office of the same 
company that happens
to have 60 Hz power. The spur moves to 60 / 120 / 180 and they move on 
with their

evaluation.

I have also played the “let’s do it with batteries" approach. Screen rooms 
were of little benefit.
Simply running test gear on batteries did not do the job. Ultimately we 
wound up in the
middle of an Illinois corn field with a bunch of gear modified to run 
purely on batteries. The
spur did go down, but it never fully went away.  It’s been 40 years since 
that adventure so I

really could not say just how far down we got it ….

Anything past -120 db at line frequency is in the “don’t bother” category.

Bob



On Jul 12, 2016, at 6:44 AM, Martyn Smith  wrote:

Hello,

I have a customer who is measuring the phase noise of my 10 MHz ultra-low 
phase noise frequency standard.


He is seeing spurious signals at line frequencies (50 and 100 Hz as we 
are in Europe) at a level around -130 dBc.


My opinion is that it's impossible to get much better than that.  Even 
running on batteries make little difference, since the equipment is in a 
test rack with AC signals everywhere.


Even the £50k R test set he is using only quotes a spurious spec of -90 
dBc.


What experience does anyone have here?

Best 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. 


___
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] OXCO Spurious Output at Line Frequencies

[jimlux]

On 7/12/16 3:44 AM, Martyn Smith wrote:

Hello,

I have a customer who is measuring the phase noise of my 10 MHz ultra-low phase 
noise frequency standard.

He is seeing spurious signals at line frequencies (50 and 100 Hz as we are in 
Europe) at a level around -130 dBc.

My opinion is that it's impossible to get much better than that.  Even running 
on batteries make little difference, since the equipment is in a test rack with 
AC signals everywhere.

Even the £50k R test set he is using only quotes a spurious spec of -90 dBc.

What experience does anyone have here?



It's impossible to get rid of that.  You'd need a magnetically shielded 
enclosure, because most of the line frequency is in the form of the 
magnetic fields from the building wiring, transformers, etc..  It's 
almost impossible to build your oscillator and measuring system without 
any loop area, into which those magnetic fields couple.


Let's assume you're at 0dBm which is 0.223 Vrms (It's easier for me to 
work in voltage, although for magnetic field, you should be working in 
current).


So, 130 dB down is 3E-7 times less, or about 74 nV.  It doesn't take a 
very big field to induce a few nV  in a circuit, even if it's all low 
impedance stuff.





___
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] OXCO Spurious Output at Line Frequencies

[Bob Camp]

Hi

The “practical” solution often is to ship it to an office of the same company 
that happens
to have 60 Hz power. The spur moves to 60 / 120 / 180 and they move on with 
their
evaluation. 

I have also played the “let’s do it with batteries" approach. Screen rooms were 
of little benefit.
Simply running test gear on batteries did not do the job. Ultimately we wound 
up in the 
middle of an Illinois corn field with a bunch of gear modified to run purely on 
batteries. The
spur did go down, but it never fully went away.  It’s been 40 years since that 
adventure so I 
really could not say just how far down we got it …. 

Anything past -120 db at line frequency is in the “don’t bother” category. 

Bob


> On Jul 12, 2016, at 6:44 AM, Martyn Smith  wrote:
> 
> Hello,
> 
> I have a customer who is measuring the phase noise of my 10 MHz ultra-low 
> phase noise frequency standard.
> 
> He is seeing spurious signals at line frequencies (50 and 100 Hz as we are in 
> Europe) at a level around -130 dBc.
> 
> My opinion is that it's impossible to get much better than that.  Even 
> running on batteries make little difference, since the equipment is in a test 
> rack with AC signals everywhere.
> 
> Even the £50k R test set he is using only quotes a spurious spec of -90 dBc.
> 
> What experience does anyone have here?
> 
> Best 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] GPS for Nixie Clock

[Vlad]

DHCP is handy to provide the IP address and specify which NTP server to 
use. Using DHCP I don't need to setup all the network settings for my 
project. The NTP packet parser itself is really small and doing its duty 
very well. Using such setup, it gives me flexibility to choose the 
different NTP servers as a time source. It could be my own NTP or it 
could be NTP pool.



On 2016-07-12 16:21, Hal Murray wrote:

t...@patoka.org said:
In addition, even MCU has not enough resources to handle TCP/IP, DHCP  
and
NTP, it is some solutions available to outsource it to dedicated  
chips. I

was using WIZ5100 (assembled as a modules) with great success.


NTP is pretty simple.  If you are willing to take a few shortcuts,
things get even simpler.  If you have a NTP server on your LAN, you
can skip routing and DHCP by broadcasting a request.  It would be
interesting to see how small you could make a set-the-time package.

The Xerox Alto had an Ethernet boot loader in microcode.  (and space
in the microcode ROM was tight)  When bringing up a new machine, we
figured out that it was much simpler and faster to debug the Ethernet
before the disk controller.  The debugger worked over the Ethernet.
That gave you a solid place to stand early in the debugging process.


--
WBW,

V.P.
___
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] GPS for Nixie Clock

[Hal Murray]

t...@patoka.org said:
> In addition, even MCU has not enough resources to handle TCP/IP, DHCP  and
> NTP, it is some solutions available to outsource it to dedicated  chips. I
> was using WIZ5100 (assembled as a modules) with great success. 

NTP is pretty simple.  If you are willing to take a few shortcuts, things get 
even simpler.  If you have a NTP server on your LAN, you can skip routing and 
DHCP by broadcasting a request.  It would be interesting to see how small you 
could make a set-the-time package.

The Xerox Alto had an Ethernet boot loader in microcode.  (and space in the 
microcode ROM was tight)  When bringing up a new machine, we figured out that 
it was much simpler and faster to debug the Ethernet before the disk 
controller.  The debugger worked over the Ethernet.  That gave you a solid 
place to stand early in the debugging process.


-- 
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] OXCO Spurious Output at Line Frequencies

[Wes]

I retired in 1999 so it's been a long time since I used one, but I don't recall 
ever seeing a device measured using an HP 3048 system that didn't exhibit line 
frequency (and harmonic) spurs.


See the "live" plots at the end of this page: 
http://hpmemoryproject.org/news/3048/hp3048_01.htm


Wes

On 7/12/2016 3:44 AM, Martyn Smith wrote:

Hello,

I have a customer who is measuring the phase noise of my 10 MHz ultra-low phase 
noise frequency standard.

He is seeing spurious signals at line frequencies (50 and 100 Hz as we are in 
Europe) at a level around -130 dBc.

My opinion is that it's impossible to get much better than that.  Even running 
on batteries make little difference, since the equipment is in a test rack with 
AC signals everywhere.

Even the £50k R test set he is using only quotes a spurious spec of -90 dBc.

What experience does anyone have here?

Best 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.

Re: [time-nuts] GPS for Nixie Clock

[Clint Jay]

I've used the wiznet module,  these days I'd consider the esp8266,
powerful little module that can be had for a couple of dollars complete
with WiFi capabilities and enough GPIO to interface GPS and some sort of
serial nixie interface.
On 12 Jul 2016 20:11, "Vlad"  wrote:

>
> In addition, even MCU has not enough resources to handle TCP/IP, DHCP and
> NTP, it is some solutions available to outsource it to dedicated chips. I
> was using WIZ5100 (assembled as a modules) with great success.
>
> Regards,
> Vlad
>
> On 2016-07-12 12:01, Chris Albertson wrote:
>
>> What kind of micro processor are you using to run the Nixie tube clock?
>> If
>> that processor could run NTP you would not need GPS.   But of course you
>> could always do both.   GPS requires a view of the sky and maybe that is
>> not available in a multi story building unless near a window.  NTP would
>> be
>> available any place there is WiFi. Even without network connection,
>> NTP
>> is a good way to interface the GPS as it knows how to talk to most GPSes
>> made. can handle GPS outages (like when the clock is moved away from a
>> window)
>>
>> That said, you are likely not using a uP big enough to run NTP as a setup
>> like that is about $40 vs. using a bare AVR chip for about $3.  But even
>> with the smaller uP you might think about having the chip keep it's own
>> internal time and using GPS to discipline that internal time, much like
>> the
>> way NTP works.   Basically the uP has a flywheel and GPS regulates it's
>> speed.  Let's you handle holdover gracefully.
>>
>> If using NMEA sentences from GPS, remember that the NMEA standard allows
>> those senates to come out at any time during the second to which they
>> apply.  In other words the sentence itself can be up to almost a second
>> "off".
>>
>> If you are looking for a GPS for use indoors I think you don't care about
>> anything other then receiver sensitivity.  Without that you have nothing,
>> no signal.  It is more important than a few less nanoseconds of
>> uncertainty
>> in the time solution.   So those ublox receivers look good.   I'm looking
>> to buy some for another application, mobile robots, I'll use GPS for gross
>> level navigation and it would be nice if it still worked indoors
>>
>> On Sun, Jul 10, 2016 at 10:24 PM, Mark Sims  wrote:
>>
>> The Jupiter receiver defaults to Motorola output, but can be switched to
>>> Zodiac.  It talks at 9600:8:N:1
>>> Frankly, it is the wrong receiver to use, particularly with an indoor
>>> antenna.  I would go with a modern GPS receiver with standard NMEA output
>>> and a 1PPS signal.   They are MUCH more sensitive and usually work
>>> indoors
>>> and can be had for dirt cheap.   Most have an on-board ceramic patch
>>> antenna.  Ublox receivers seem to work well indoors.
>>>
>>> NAVSPARK makes a tiny little GPS board with 1PPS output.  6 for $36 or
>>> one
>>> free for $10 shipping (no antenna supplied, has a U.FL connector).  It
>>> speaks NMEA at 115,200:8:N:1  or can be setup for 9600 baud.
>>>
>>> The delay between message and 1PPS is receiver dependent.  Usually it is
>>> small enough to not be noticeable.  Some receivers send the message
>>> before
>>> the 1PPS, others after it.   You can compensate for the
>>> differences/delays
>>> in software.
>>>
>>>
>>> ___
>>> 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.
>>>
>>>
> --
> WBW,
>
> V.P.
> ___
> 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] GPS for Nixie Clock

[Vlad]

In addition, even MCU has not enough resources to handle TCP/IP, DHCP 
and NTP, it is some solutions available to outsource it to dedicated 
chips. I was using WIZ5100 (assembled as a modules) with great success.


Regards,
Vlad

On 2016-07-12 12:01, Chris Albertson wrote:
What kind of micro processor are you using to run the Nixie tube clock? 
 If
that processor could run NTP you would not need GPS.   But of course 
you
could always do both.   GPS requires a view of the sky and maybe that 
is
not available in a multi story building unless near a window.  NTP 
would be
available any place there is WiFi. Even without network connection, 
NTP
is a good way to interface the GPS as it knows how to talk to most 
GPSes

made. can handle GPS outages (like when the clock is moved away from a
window)

That said, you are likely not using a uP big enough to run NTP as a 
setup
like that is about $40 vs. using a bare AVR chip for about $3.  But 
even

with the smaller uP you might think about having the chip keep it's own
internal time and using GPS to discipline that internal time, much like 
the

way NTP works.   Basically the uP has a flywheel and GPS regulates it's
speed.  Let's you handle holdover gracefully.

If using NMEA sentences from GPS, remember that the NMEA standard 
allows

those senates to come out at any time during the second to which they
apply.  In other words the sentence itself can be up to almost a second
"off".

If you are looking for a GPS for use indoors I think you don't care 
about
anything other then receiver sensitivity.  Without that you have 
nothing,
no signal.  It is more important than a few less nanoseconds of 
uncertainty
in the time solution.   So those ublox receivers look good.   I'm 
looking
to buy some for another application, mobile robots, I'll use GPS for 
gross

level navigation and it would be nice if it still worked indoors

On Sun, Jul 10, 2016 at 10:24 PM, Mark Sims  wrote:

The Jupiter receiver defaults to Motorola output, but can be switched 
to

Zodiac.  It talks at 9600:8:N:1
Frankly, it is the wrong receiver to use, particularly with an indoor
antenna.  I would go with a modern GPS receiver with standard NMEA 
output
and a 1PPS signal.   They are MUCH more sensitive and usually work 
indoors

and can be had for dirt cheap.   Most have an on-board ceramic patch
antenna.  Ublox receivers seem to work well indoors.

NAVSPARK makes a tiny little GPS board with 1PPS output.  6 for $36 or 
one

free for $10 shipping (no antenna supplied, has a U.FL connector).  It
speaks NMEA at 115,200:8:N:1  or can be setup for 9600 baud.

The delay between message and 1PPS is receiver dependent.  Usually it 
is
small enough to not be noticeable.  Some receivers send the message 
before
the 1PPS, others after it.   You can compensate for the 
differences/delays

in software.


___
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.



--
WBW,

V.P.
___
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] OXCO Spurious Output at Line Frequencies

[Scott Stobbe]

Unless your DUT is in a mu-metal enclosure, I wouldn't say observing the
same results while battery powered provides any conclusive statement as to
whether or not its the DUT. The terminals around a crystal are pretty high
impedance at line frequencies.

On Tue, Jul 12, 2016 at 6:44 AM, Martyn Smith  wrote:

> Hello,
>
> I have a customer who is measuring the phase noise of my 10 MHz ultra-low
> phase noise frequency standard.
>
> He is seeing spurious signals at line frequencies (50 and 100 Hz as we are
> in Europe) at a level around -130 dBc.
>
> My opinion is that it's impossible to get much better than that.  Even
> running on batteries make little difference, since the equipment is in a
> test rack with AC signals everywhere.
>
> Even the £50k R test set he is using only quotes a spurious spec of -90
> dBc.
>
> What experience does anyone have here?
>
> Best 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.

[time-nuts] More FE405B data

[cdelect]

Hi,

I posted an AD plot of a good FE405B back on Jun 7th.

A plot just completed today after an additional month of running is
essentially the same.

I did however measure the aging rate and it is currently 1.15X10-11th per
day!

Cheers,

Corby

___
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] A different way to think about time dilation?

[Chris Albertson]

Using Google I find this idea is all over theInternet and quite common.
That motion in 4-space is always at c.  Or more specifically motion in four
space for a body at rest in 3-space is v =  (dx, dy, dz, dt/t) = (0, 0, 0,
c)  for a photon trading along x axis it is v = (c, 0, 0, 0)  and that |(c,
0, 0, 0) | = |(0, 0, 0, c) |

When I search discussions there seem to be to camps, those who say "of
course, it is implied by special relativity" and those who say "that is not
the way my old text books presented it, so it must be wrong"

I'm becoming a member of the first camp just by common sense.  In 4-space
there can be no such thing as a stationary body.  It is alway moving in
time.  OK not always, if it is moving through space at c it is not moving
in time but it is not stationary.

You can find any number of derivations using Google (as well as any number
of people arguing that the concept on "moving" in 4-space makes no sense.)



On Mon, Jul 11, 2016 at 12:21 PM, William H. Fite  wrote:

> Chris, if you want to verify the mathematical operations you can send them
> to me and my spouse and I will check them for you. As to whether you chose
> the right equations, you probably need the help of a physicist.
>
> Bill
>
> On Monday, July 11, 2016, Chris Albertson 
> wrote:
>
> > What I really asked was "does the math work?".  So far I suspect it does.
> > I don't think what I wrote contradicts anything in any conventional text
> > book.  What I'm looking for is to be proven wrong
> >
> > Yes I know about velocity driven time dilation.  Let's stick with Special
> > Relativity for now and ignore gravity.   Notice that in this alternate
> > explanation thinks work the same way.  It they don't then I'm proven
> > wrong.  The way to prove me wrong is to compute the same situation both
> > ways and get different answers in just one case (that is not some special
> > corner case)
> >
> > Notice that your use of "velocity" or speed is confined to only 3-space.
> > Notice in my different explanation when speed in x,y,z is zero time is
> > moving at a 1:1 ratio and when speed in x,y,z is equal to c then time is
> > moving at zero speed.   Al I did was ask what happens if we talk about
> > speed in x,y,z,t or "4-space". My first guess is that it would make
> > everything so complex no one would want to think about it but no, it
> seems
> > to make it easier because you only need to think about a plane parallel
> to
> > t axis, no need to think in 4-space, 2-space is general enough
> >
> > So I'm certainly NOT challenging anything in Special Relativity.  I've
> read
> > what Einstein has written on this and I think all his examples apply
> What
> > you wrote is true also.   You are using Einstein's examples. They are
> > good.  But he and you are talking about speed in 3-space.
> >
> > I think it is intuitive that I am right now not moving in x,y,z but I
> KNOW
> > I am moving in "t" (time) at about 1 second/secind and from my reference
> > point I NEVER MOVE I am always "here" so I always experience time at 1
> s/s
> >  So I forgot to say that the x,y,z,t frame is relative to some "fixed"
> > object like my office.We all know that we are moving in time even if
> we
> > have no control over it.  If we are moving then we should be able to
> > measure our velocity.  Velocity is always something over time.   It this
> > case it must be time over time.  Using units it becomes seconds per
> second.
> >   Then you set 1 s/s = c (tally arbitrary assignment) and much complexity
> > falls out.
> >
> > No intention to invent new physics here, just a different way to compute
> > and explain the same thing.   It works the same way an observer in my
> > office sees me push my chair back at 4 inch/second and sees that my watch
> > has slowed down by some tiny amount.   I claim only that assuming every
> > object in the universe always moves in 4-space at speed = c  makes the
> > calculation simpler and easier to understand.
> >
> >
> >
> > On Sun, Jul 10, 2016 at 9:30 PM, Bill Byrom  > > wrote:
> >
> > > I think you are on the wrong track with assuming that every object has
> a
> > > velocity c. What you need to consider is relativity. Velocity is a
> local
> > > measurement (local reference frame distance and local reference frame
> > > time). Light (and other electromagnetic radiation) always travels at a
> > > local velocity c (local distance divided by local time). Time dilation
> > > is a way of describing the effect of the relativity of simultaneity.
> > > Events which are not local (adjacent) to each other can't be
> > > unambiguously described as simultaneous. There is no universal clock
> > > which allows us to determine which of two separated events occurred
> > > "before" the other.
> > >
> > > There are two causes of time dilation:
> > > (1) Relative uniform motion. If two spacecraft are passing each other
> > > in uniform motion 

Re: [time-nuts] GPS for Nixie Clock

[Chris Albertson]

What kind of micro processor are you using to run the Nixie tube clock?  If
that processor could run NTP you would not need GPS.   But of course you
could always do both.   GPS requires a view of the sky and maybe that is
not available in a multi story building unless near a window.  NTP would be
available any place there is WiFi. Even without network connection, NTP
is a good way to interface the GPS as it knows how to talk to most GPSes
made. can handle GPS outages (like when the clock is moved away from a
window)

That said, you are likely not using a uP big enough to run NTP as a setup
like that is about $40 vs. using a bare AVR chip for about $3.  But even
with the smaller uP you might think about having the chip keep it's own
internal time and using GPS to discipline that internal time, much like the
way NTP works.   Basically the uP has a flywheel and GPS regulates it's
speed.  Let's you handle holdover gracefully.

If using NMEA sentences from GPS, remember that the NMEA standard allows
those senates to come out at any time during the second to which they
apply.  In other words the sentence itself can be up to almost a second
"off".

If you are looking for a GPS for use indoors I think you don't care about
anything other then receiver sensitivity.  Without that you have nothing,
no signal.  It is more important than a few less nanoseconds of uncertainty
in the time solution.   So those ublox receivers look good.   I'm looking
to buy some for another application, mobile robots, I'll use GPS for gross
level navigation and it would be nice if it still worked indoors

On Sun, Jul 10, 2016 at 10:24 PM, Mark Sims  wrote:

> The Jupiter receiver defaults to Motorola output, but can be switched to
> Zodiac.  It talks at 9600:8:N:1
> Frankly, it is the wrong receiver to use, particularly with an indoor
> antenna.  I would go with a modern GPS receiver with standard NMEA output
> and a 1PPS signal.   They are MUCH more sensitive and usually work indoors
> and can be had for dirt cheap.   Most have an on-board ceramic patch
> antenna.  Ublox receivers seem to work well indoors.
>
> NAVSPARK makes a tiny little GPS board with 1PPS output.  6 for $36 or one
> free for $10 shipping (no antenna supplied, has a U.FL connector).  It
> speaks NMEA at 115,200:8:N:1  or can be setup for 9600 baud.
>
> The delay between message and 1PPS is receiver dependent.  Usually it is
> small enough to not be noticeable.  Some receivers send the message before
> the 1PPS, others after it.   You can compensate for the differences/delays
> in software.
>
>
> ___
> 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.

Re: [time-nuts] OXCO Spurious Output at Line Frequencies

[Stéphane Rey]

Hi Martyn,

For such measurement you may shield your DUT and instruments. Ever using 
a shielded room which might not be easy for everyone or using shielded 
tents. You can find them made custom for 600-1000€ depending on the 
size. I've just ordered two of them from a company in netherlands


For main line frequencies, you may differentiate what is conducted from 
what is radiated. For EMI, I'm using now active low pass filter at very 
low frequency to follow my low noise regulators and I get very good 
result, but honestly not tested as low as -130dBc... But on low noise 
PLL or oscillators I can get rid of any spurious from PSU but at 50Hz, 
my noise floor has never reached -130dBc...  I'm working on a very low 
noise generator (20fs jitter 10Hz-1Mhz) at the moment and at 50 Hz, the 
phase noise is  -80dBc/Hz which is already not bat at 6 GHz


I've measured last week a Wentzel VCXO giving -150dBc/Hz @ 50 Hz and I 
could measure the 50 Hz at -130dBc/Hz but this had not the active filter 
on it. I test it again using the shielded tent exactly to know what is 
radiated from conducted. If still there under the tent I will try using 
my active filter to see if this makes a difference.



Cheers
Stephane

-- Message d'origine --
De : "Martyn Smith" 
À : time-nuts@febo.com
Envoyé 12/07/2016 12:44:31
Objet : [time-nuts] OXCO Spurious Output at Line Frequencies


Hello,

I have a customer who is measuring the phase noise of my 10 MHz 
ultra-low phase noise frequency standard.


He is seeing spurious signals at line frequencies (50 and 100 Hz as we 
are in Europe) at a level around -130 dBc.


My opinion is that it's impossible to get much better than that.  Even 
running on batteries make little difference, since the equipment is in 
a test rack with AC signals everywhere.


Even the £50k R test set he is using only quotes a spurious spec of 
-90 dBc.


What experience does anyone have here?

Best 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] OXCO Spurious Output at Line Frequencies

[timeok]

Hi,

I agree with you. Some HP spectrum analyzer have  internal 50 and 100 Hz spurs 
down to -80dBc due the fan as described in the manual. It can be the spurs are 
from the test set and/or connection cables. If your oscillator is battery 
powered you can be sure the spurs are not from the source. Anyway I think the 
-130 dBc value is a very good response.

Luciano
www.timeok.it


From: "time-nuts" time-nuts-boun...@febo.com
To: time-nuts@febo.com
Cc: 
Date: Tue, 12 Jul 2016 11:44:31 +0100
Subject: [time-nuts] OXCO Spurious Output at Line Frequencies


Hello,
 
I have a customer who is measuring the phase noise of my 10 MHz ultra-low phase 
noise frequency standard.
 
He is seeing spurious signals at line frequencies (50 and 100 Hz as we are in 
Europe) at a level around -130 dBc.
 
My opinion is that it's impossible to get much better than that.  Even running 
on batteries make little difference, since the equipment is in a test rack with 
AC signals everywhere.
 
Even the £50k R test set he is using only quotes a spurious spec of -90 dBc.
 
What experience does anyone have here?
 
Best 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.

[time-nuts] OXCO Spurious Output at Line Frequencies

[Martyn Smith]

Hello,

I have a customer who is measuring the phase noise of my 10 MHz ultra-low phase 
noise frequency standard.

He is seeing spurious signals at line frequencies (50 and 100 Hz as we are in 
Europe) at a level around -130 dBc.

My opinion is that it's impossible to get much better than that.  Even running 
on batteries make little difference, since the equipment is in a test rack with 
AC signals everywhere.

Even the £50k R test set he is using only quotes a spurious spec of -90 dBc.

What experience does anyone have here?

Best 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.