DIY Frequency Standard - crystal?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Freq Band

Well-known member
GDIY Supporter
Joined
Jan 5, 2006
Messages
608
Location
Electra City
Would it be worthwhile (and is it feasible) to build my own freq standard based around a crystal oscillator ?
This particular 10MHz part is a military "pull" , and the seller says it has been checked :

10mhzvec.jpg


(hook-up diagram is included....15v)

I need a reliable reference to align some old test equipment.


thanks-
=FB=
 
What are your exact requirements w.r.t. functionality & performance ?

In other words, do you need to 'arrive' at audio-stuff frequencies 44k1, 96k etc (then better starting from another freq for easy of 'dividing by 2'), or is it for adjusting/use with lab/test-equipment ? (then the 10MHz can indeed come in handy since a 'common' frequency).

Also note that however cool this thing looks it may not be fully accurate anymore... and how to determine that without some other stuff around ?


BTW, is it one of those octal-type assemblies ? Then rip the guts and build an Altec-type preamp in it :wink:
 
Yes, 10MHz is the freq stated in my test equipment manuals as what's needed to calibrate....or use as an external reference.

...but you are right, how do I check this reference :? ?
(as in...which came first , chicken or egg?)

The seller "says" it has been checked already.


=FB=
 
[quote author="Freq Band"]Yes, 10MHz is the freq stated in my test equipment manuals as what's needed to calibrate....or use as an external reference.

...but you are right, how do I check this reference :? ?
(as in...which came first , chicken or egg?)

The seller "says" it has been checked already.


=FB=[/quote]

Right, chicken & egg :wink:

What's the complete module doing ? A complete oscillator, probably complete with oven ? Or just the Xtal & oven ? Obviously a simple CMOS-inverter and a few caps can make already a suited test-circuit, but will the seller do that for you ? So any idea how he has checked it ?

What's the price BTW ? If it's cheap then just buy & try to find someone nearby with a calibrated counter. There may be other tricks perhaps, like comparing with a simple but easily available frequency, dunno.

(FWIW: I recently plugged a gtr both into a Boss gtr-tuner and an HP-THD-test set and both were in very good agreement w.r.t. measured frequency... which tells me that while both units haven't been calibrated in years, it's unlikely that they're both really wrong, since then they had to be off in exactly the same manner, which is unlikely so most likely they're both pretty accurate... :thumb: )

So what's the accuracy you need ? Any reason to assume your lab-stuff is 'off' ? And if that's indeed the case, how bad is it that you've determined the range of an LFO a few % off ?

Bye,

Peter
 
Here is a 10 MHz oscillator schematic (w/fet buffer). What layout considerations do you think are necessary ?
(PostScript file)
http://www.wireless-homebrew.com/wireless/ps/10mhz.ps
(I'm not sure who to give credit for this schemo, but it's from 45SMJ:www.wireless-homebrew.com)
EDIT: here it is
10MHzStandard.gif


---------------

...and another Standard, derived from a satellite broadcasting a 1 second pulse:
http://www.rt66.com/~shera/QST_GPS.pdf
(too involved for me though)

---

Here is the link to the seller, of the part I mentioned:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ih=008&sspagename=STRK%3AMEWA%3AIT&viewitem=&item=180134322338&rd=1&rd=1

=FB=
 
The time transmission from the NTSI is on multiples of 5MHz. This frequency is controlled by the atomic standard. You need a shortwave receiver to listen to the heterodyne between the primary standard frequency and your local oscillator. You may need to attenuate your local oscillator output in order to do this, depending on how close you are to Boulder. You adjust your local oscillator to minimize the beat.
 
[quote author="Freq Band"]
(I'm not sure who to give credit for this schemo,[/quote]
Mr. Pierce :thumb:


edit: wait, I got fooled by the connection to the 470 resistor. OK, so the 10MHz is a complete circuit of its own, not just a Xtal-like thingy.

While this item sure looks a lot more funky I'm still wondering if your applications are perhaps equally well served by a simple CMOS-gate, 10.000MHz Xtal, a few caps and a buffer. Just a sanity check.


Regards,

Peter
 
Like clintrubber asked earlier, what do you want to use this for ? You mention alignment, what do you need to align ?

Any idea on the accuracy you need ? 1% ? 0.1% ? 10ppm ? <1ppm ?

Do you care about jitter / phase noise ?

What kind of voltages do you need ? Is this TTL, or RF ?

JDB.
 
The freq is reletive easy to check.

WWV transmits on 10 MHz so get a shortwave receiver and tune to 10 mHz. depending on the time of the day (atmospheric effects) you should hear the time code ticking away. Turn on your osc and you should hear a beat frequency. You can look at the beat freq and measure how accurate your osc is.

Based on the age of the osc I presume that it is a sine wave out. Of course if your osc is too far offset from 10 mHz you could have a beat freq beyond the audio range....

After letting it warm up you can adjust it for a null.
 
quibble: Used as prefixes, convention says M stands for mega; m for milli.

Some simulators are not case-sensitive and interpret all m's as milli. Circuitmaker requires that "Meg" or "meg" be used, for example.
 
[quote author="bcarso"]quibble: Used as prefixes, convention says M stands for mega; m for milli.

Some simulators are not case-sensitive and interpret all m's as milli. Circuitmaker requires that "Meg" or "meg" be used, for example.[/quote]
More spelling-police here: in that particular message I spotted a dot after 'mHz' !

Hz is a symbol, not an abbrev. :wink:

Likewise: 'sec.'.... --> just 's'

Enough for now, back to ppm :thumb:
 
LOL

The small m is a typo in both places. You will note the MHz in the initial statement. As for the period after the mHz (which should be MHz), it is correct. The "d" of the next word was not a capitol as it should of been as the starting word of a statement.

So we see here that the spelling police missed the point....

I should note that I am exceeding bad at spelling and tend not to worry too much about correcting typos Somehow I knew that I would be called on the mHz but being I'm at work and have 5 things up in the air I did not take the time to correct it.

Happy 4th!
 
[quote author="jdbakker"]Like clintrubber asked earlier, what do you want to use this for ? You mention alignment, what do you need to align ?
JDB.[/quote]
Keithley 776 freq counter
and check my Tek CFG280 funct. gen.

[quote author="jdbakker"]Any idea on the accuracy you need ? 1% ? 0.1% ? 10ppm ? <1ppm ?[/quote]
The Keithley counter has 9 digit resolution :shock: I don't haven't received the unit yet, but the ebay seller could not get a correct reading. The 10MHz reference output on the rear, when connected to the front input, gave this reading:
http://www.bpai.us/min/KEITHLEY_776-3.jpg
(It's possible he does not have it setup correctly. There is also an internal jumper, designating weather or not to use it's internal 10MHz, or an external 10MHz reference.

[quote author="jdbakker"]Do you care about jitter / phase noise ?[/quote]
should I ? XTAL/TCXO's and HF are new territory for me.

----------------

[quote author="evm1024"] WWV transmits on 10 MHz so get a shortwave receiver and tune to 10 mHz............
.......... Turn on your osc and you should hear a beat frequency. You can look at the beat freq and measure how accurate your osc is.[/quote]
I have a cheap one, a "Realistic DX-150A" with a 100 foot wire stretched accross my yard.
...but I also have a beautiful big'old Philco "frequency selective voltmeter" that I've never used...I' haven't messed with it yet. I have the manual.

------------------------------------

[quote author="burdij"]The time transmission from the NTSI is on multiples of 5MHz......... You need a shortwave receiver to listen to the heterodyne between the primary standard frequency and your local oscillator. You may need to attenuate your local oscillator output in order to do this, depending on how close you are to Boulder. You adjust your local oscillator to minimize the beat.[/quote]
"minimizing the beat"....is this like a canceling-out routine....if they are in sync, they cancel each other? (or "step on" each other?)

----------------------------------------------
Thanks for your replies.

your willing student,
=FB=
 
[quote author="evm1024"]LOL
... Somehow I knew that I would be called on the mHz but being I'm at work and have 5 things up in the air I did not take the time to correct it.

Happy 4th![/quote]

Thanks for having a sense of humor about it. I just worry about impressionable young minds---I guess it's a generational thing. :razz:
 
Today I purchased two HP quartz oscillators made in the mid 1960's....
103AR,
and a
101A.
Both put out a 100 KC, and 1 MC sine signal.
I found the manual to the 103AR on-line, but it needs a seperate Power supply that has battery backup (HP 724BR or 725AR). It takes up to 3 weeks to temperature stabilize, as it has an oven within an oven. I am trying to research if I can build a PSU, or if I can just test it with my lab psu. It takes -24v, 500ma.....I think, because those are the only psu specs I could find.
DSCN2342.jpg

http://i5.photobucket.com/albums/y177/Midiot/DSCN2343.jpg
http://i5.photobucket.com/albums/y177/Midiot/DSCN2344.jpg

Sorry if this is a bit off topic, it's not very audio related.

=FB=
 

Latest posts

Back
Top