Hi Euclides,

On 02/08/13 18:31, Euclides Chuma wrote:

There are some benefits (traditionally) in using 5 MHz over 10 MHz, but
10 MHz have become a common standard. The actual frequency isn't really
magic, but 5 MHz and multiples became somewhat standard in the old MIL
STD 188 for time-keeping, and it fit fairly well with what was already
in use. There are folks here that can correct me on massive details.

Today 10 MHz is more common because, well, engineers then to be
following habits, and 10 MHz "sounds nice". I use 10 MHz mainly because
the application requires it, otherwise I use whatever frequency fits my
other needs, or what becomes easy to source.

PS. Have not seen you post before, so welcome to time-nuts!

Cheers,
Magnus

It is my understanding that Quartz has a sweet spot at 5MHz that makes it ideal if the lowest possible phase noise and highest stability are needed.

Thomas Knox

Many important lab references over past 50, 60 years are named and
characterized at http://www.ieee-uffc.org/main/history-norton.asp

2.5MHz and 5 MHz seem common as far back as the 1950's and 1960's (and I've
used some of them! e.g. Sulzer) and continue through today.

Late 60's and early 70's HP benchtop counters used a 1 MHz crystal. e.g. HP
5321B.

By the mid-70's HP Frequency counters were using 10MHz references. e.g. Hp
5383A (which for many of us was the defacto bench counter forever) most
often in a TCXO can.

Tim.

On Fri, Aug 2, 2013 at 12:31 PM, Euclides Chuma euclides@w2c.com.br wrote:

Hi

If you go back far enough, you will find 1 MHz and 100 KHz used as reference standards. They certainly were common place in gear through the 1960's.

Pre-WWII,  open 100 Kc quartz bars were often the "standard" of choice. Anything much higher than that was increasingly less stable.

Post WWII,  well behaved 1  Mc  fundamentals came along in glass sealed enclosures. Sealing them up helped a lot with aging. By the late 40's and early 50's low frequency overtone designs (1 to 5 MHz) were worked out.

5 MHz third overtones (and 5th's) became popular with the Bell Labs / Western Electric / AT&T crowd by the mid 1950's. They made and bought enough of them to turn them into a standard by the early 1960's. As cold weld packages took over from glass, the older 2.5 and 1 MHz 3rd's / 5th's became impractical.

10 MHz is a creation mainly of HP. They were more after the microwave end of the business than some of the others. The higher starting frequency gave them better phase noise and fewer spur issues.

Depending on where you are in the world, 10 MHz may or may not have caught on. The Russians stuck with 5 MHz long after HP had converted the US over to 10.

Intrinsically quartz just keeps getting better as you go down in frequency. There is nothing magic about any one frequency. Lower is always better in terms of Q, provided you can scale up all the dimensions. A 1 MHz OCXO that is 10X the outer dimensions of an 10 MHz OCXO obviously would have some limitations that had noting to do with quartz. As an example, the blank in a 5MHz precision crystal is probably 0.5" diameter. To make an equivalent 1 MHz crystal, the blank would need to be 2.5" in diameter. At 100 KHz you would have a plate 25" in diameter. Good luck finding the raw quartz for those blanks ….

Bob

On Aug 2, 2013, at 12:31 PM, Euclides Chuma euclides@w2c.com.br wrote:

It was my understanding that this "sweet spot" was optimum a little above 3
MHz, so, 3rd overtone crystals are used to generate a stable, low phase
noise 10 MHz.  Prior to that, 5 MHz was used and before that 1 MHz  Regards

• Mike

Mike B. Feher, EOZ Inc.
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960 office
908-902-3831 cell

-----Original Message-----
From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com] On
Behalf Of Tom Knox
Sent: Friday, August 02, 2013 2:02 PM
To: Time-Nuts
Subject: Re: [time-nuts] 5MHz x 10MHz

It is my understanding that Quartz has a sweet spot at 5MHz that makes it
ideal if the lowest possible phase noise and highest stability are needed.

Thomas Knox

details.

time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Hi

Quartz it's self has no "sweet spot". The only issue is how low you can go in a specific sized crystal holder before you start to run into trouble. A TO-5 crystal will have a different minimum frequency than an HC-40.

Bob

On Aug 2, 2013, at 2:30 PM, Mike Feher mfeher@eozinc.com wrote:

Hi

It may well have been teamed up with a piece of Russian designed equipment.

Bob

On Aug 2, 2013, at 2:40 PM, Euclides Chuma euclides@w2c.com.br wrote:

Hi
Most of the British Racal standards are 5MHz. It may well have been down to what was the best performance of the nationally avilable crystals. Everthing is a compromise.
It is easy to double a 5MHz output to 10MHz. One way is to pass it through a bridge rectifier (high speed diodes of course) and then filter it. old 10Mbs "thin" ethernet filters from network cards work well. Check the archives and do a websearch.

Robert G8RPI.

From: Bob Camp lists@rtty.us
To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Friday, 2 August 2013, 20:12
Subject: Re: [time-nuts] 5MHz x 10MHz

Hi

It may well have been teamed up with a piece of Russian designed equipment.

Bob

On Aug 2, 2013, at 2:40 PM, Euclides Chuma euclides@w2c.com.br wrote:

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Hi

You might want to check the equipment you are going to use with it. A lot of early gear that is marked "10 MHz" really has a range of frequencies it will accept. They often will accept anything that is a sub-multiple of 10 MHz (5, 2.5, 3.33333, 1.25 etc). They used a simple phase detector to lock up a local 50 or 100 MHz oscillator in the equipment.

Bob

On Aug 2, 2013, at 2:40 PM, Euclides Chuma euclides@w2c.com.br wrote:

Hi

Even today you will get better ADEV off of a large package (HC-40) 5 MHz crystal than off of a 10 MHz device. Of course very few people make such an oscillator any more. Buyers are after other things.

Bob

On Aug 2, 2013, at 3:51 PM, Robert Atkinson robert8rpi@yahoo.co.uk wrote:

Hi

Except for HP, everything from the US would have been 5 MHz as well.

Bob

On Aug 2, 2013, at 3:51 PM, Robert Atkinson robert8rpi@yahoo.co.uk wrote:

Interesting, I have heard for years from the senior Time and Freq researchers I work with that 5MHz was a sweet spot. I will ask if there is a reason and proven physics behind it but these are individuals that are well grounded in science.  They almost always multiply 5MHz if they needed 10MHz etc.
Perhaps I missed something. It wouldn't be the first time I was schooled by the TimeNuts.
Best Wishes;
Thomas Knox

Even HP used 5MHz when it counted (no pun intended!) - the 5245L used a 1MHz
timebase while the higher spec timebase in the 5245M used 5MHz.
DaveB, NZ

----- Original Message -----
From: "Bob Camp" lists@rtty.us
To: "Discussion of precise time and frequency measurement"
time-nuts@febo.com
Sent: Saturday, August 03, 2013 8:47 AM
Subject: Re: [time-nuts] 5MHz x 10MHz

Hi

The math is pretty simple:

The Q of quartz goes up as the frequency goes down.

A crystal resonator's performance (Q)  is limited by it's thickness to diameter ratio.

At some point the resonator design impacts the Q of the resonator more than the Q of the raw quartz.

Holders are available that will rationally hold a maximum diameter blank.

It's the intersection of all of the above that implies a best solution. The "sweet spot" is not just quartz, it's the combination of all of the above.

Change any of the above (like the holder) and you get another "sweet spot"

Is that simple? Of corse not. Many things need to change to let you make a high performance blank that's much bigger. Many things need to change to keep the Q of the quartz the limiting factor.

Why has it not been done? The drive in the marketplace is to smaller / cheaper. This is totally the opposite direction from that. The investment to make larger blanks goes at least back to the design of the gear that grows quartz. What we have is "good enough", but it's far from the best we could do. Quartz is not the limiting factor.

Bob

On Aug 2, 2013, at 7:52 PM, Tom Knox actast@hotmail.com wrote:

Some instruments (Tek 494 for instance) use a 100 MHz VCXO phase locked to 10 MHz for lower phase noise when multiplied into the microwave bands, demonstrating that there is more than one way to skin a cat.

Didier KO4BB

Bob Camp lists@rtty.us wrote:

--
Sent from my Motorola Droid Razr 4G LTE wireless tracker while I do other things.

Hi

The whole drop frequency / better Q thing really only applies if you are looking for ADEV with tau's > = 0.1 second.  If you are after phase noise, then there are other things to worry about.

Bob

On Aug 3, 2013, at 7:52 AM, Didier Juges shalimr9@gmail.com wrote:

On 08/03/2013 02:28 PM, Bob Camp wrote:

White noise and flicker noise of oscillator and buffer amps comes to
mind. Naturally noise in the crystal itself.

Cheers,
Magnus