Author Topic: OCXO calibration  (Read 13654 times)

0 Members and 1 Guest are viewing this topic.

Offline CodebirdTopic starter

  • Regular Contributor
  • *
  • Posts: 161
  • Country: gb
OCXO calibration
« on: September 16, 2017, 09:14:37 am »
I've built myself a little precision frequency reference. It uses an OCXO. Specifically, this one:

http://www.isotemp.com/wp-content/uploads/2011/10/OCXO143-3.pdf

Sure, it's less stable than a GPS disciplined oscillator, but it's less than a tenth the price and it's ready to go after just a minute to warm up. The rest of the electronics in my reference box are just a CD4020 to divide it down into convenient outputs, a 50ohm driver, voltage regulator and some level/frequency selection controls.

But, I got that reference off of eBay. And according to the datasheet, it's only guaranteed accurate to 3.5e-7 after ten years - which isn't far off the age on it now. That's still pretty good though, I think? I'm not sure if my calculations are right, but I think it comes out to about the same as most TCXOs? 0.35ppm. Unlike a TCXO though, I have the option to calibrate it.

What I'm wondering now is how I might go about calibrating such a thing. I'm in the UK, so I don't know if there are any good frequency reference stations I might be able to pick up. Any suggestions? Or anyone else in the country with access to the right equipment who would like to help? I don't know how much a laboratory would charge, and I doubt they'd be happy about touching anything hobbyist-made.

The purpose of this box is testing cheap, super-low-cost frequency counters off of eBay so that I can either be impressed with their low-cost performance or mock them for including useless extra digits, depending on results. I am curious to see how the El Cheapo bare-PCB counters compare with the thousand-pound-plus professional counters.
 

Offline MK

  • Regular Contributor
  • *
  • Posts: 233
  • Country: gb
Re: OCXO calibration
« Reply #1 on: September 16, 2017, 09:31:24 am »
You could build a long wave reciever for Rugby (60KHz) or Radio 2 longwave (Droitwich 198KHz) and divide down to 1KHz and compare the two by looking at the beat note?
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 9410
  • Country: gb
Re: OCXO calibration
« Reply #2 on: September 16, 2017, 09:55:12 am »
Buy yourself a cheap Ublox GPS module off ebay. In brief:

- Divide your OCXO down to 1MHZ.
- Set one of the Ublox module frequency outputs to 1MHz.
- Compare the two signals on a scope, ignoring any jitter, just look at the long term roll between the two traces.
- Tweak the frequency adjustment voltage of the OCXO for zero roll.
- Repeat as often as you feel necessary.

GPS module teardown here...
https://www.eevblog.com/forum/projects/ebay-u-blox-lea-6t-gps-module-teardown-and-initial-test/msg886887/#msg886887


P.S. To configure / monitor the Ublox module you use their u-center s/w... https://www.u-blox.com/en/product/u-center-windows
« Last Edit: September 16, 2017, 10:06:31 am by Gyro »
Best Regards, Chris
 

Online DimitriP

  • Super Contributor
  • ***
  • Posts: 1288
  • Country: us
  • "Best practices" are best not practiced.© Dimitri
Re: OCXO calibration
« Reply #3 on: September 16, 2017, 07:34:14 pm »
Quote
The purpose of this box is testing cheap, super-low-cost frequency counters off of eBay so that I can either be impressed with their low-cost performance or mock them for including useless extra digits, depending on results. I am curious to see how the El Cheapo bare-PCB counters compare with the thousand-pound-plus professional counters.

The "cheapo" counters don't have enough digits.
For all practical purposes, old, aged and all, you have a 10.000.000.x Hz oscillator :)


Illustrated exambles fro current ebay listings:
http://www.ebay.com/itm/High-Accuracy-RF-1-to-500-MHz-Frequency-Counter-Tester-measurement-For-ham-Radio-/281423258318?epid=1147058638&hash=item41862222ce:g:hCsAAMXQRPRTHhjB


http://www.ebay.com/itm/High-Accuracy-1-500MHz-Frequency-Counter-meter-Antenna-for-Ham-Radio-Interphone/321820986581?_trkparms=aid%3D222007%26algo%3DSIM.MBE%26ao%3D2%26asc%3D41376%26meid%3De58eba64db0d41bdaa6e36fba88e4c50%26pid%3D100005%26rk%3D6%26rkt%3D6%26sd%3D281423258318&_trksid=p2047675.c100005.m1851


You can build a cheap GPSDO, for it, but that's another rabbit hole entirely ! :)
   If three 100  Ohm resistors are connected in parallel, and in series with a 200 Ohm resistor, how many resistors do you have? 
 
The following users thanked this post: cdev

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: OCXO calibration
« Reply #4 on: September 17, 2017, 03:18:47 am »
Buy yourself a cheap Ublox GPS module off ebay. In brief:

The freq outputs of the Ublox module are derived from the (undisciplined) oscillator on the module (could be a XO or a TCXO) that is further degraded by the GPS sawtooth error and frequency divider jitter.   So what you are trying to do is to adjust a OCXO (with maybe 1E-11 stability) against a cheap TCXO with maybe 1E-8 stability / accuracy... a VERY losing proposition.  To properly set a OCXO on freq requires a GPSDO or cesium beam oscillator.  Or maybe a hight quality rubidium oscillator that was recently calibrated against GPSDO or cesium oscillator...  and all those telecom grade rubidiums are NOT particularly high quality rubidiums.   
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 2260
  • Country: ca
Re: OCXO calibration
« Reply #5 on: September 17, 2017, 03:59:45 am »
Your calibration method will depend on how close you want the frequency to be.

Note that the daily drift for your OCXO is spec'ed at 1e-9, i.e. 0.01 Hz/day.  Do you want to adjust it every day?  Probably not.  Once a year?  That would mean that the frequency will be within 1e-7 if it meets its spec., i.e. 100 PPB or 1 Hz for your 10 MHz OCXO.  As texaspyro said, GPS by itself has a stability of about 1e-8, i.e. 10 PPB for a one second measurement with no overall frequency error.  A 10 sec. gate time would give you 1 PPB.  That sounds like a good match.  A bare GPS receiver could be used as a reference.  If these numbers aren't good enough for you, you'll have to upgrade to a better OCXO, a GPSDO or a Rubidium.

You said "it's ready to go after just a minute to warm up".  No, it isn't!  The aging specs only apply if you keep it powered up all the time.  A single power failure will wreak havoc with the aging.  You'll have to try it to see how big the effect will be for your particular unit.  OCXOs typically age fast after a power cycle.  Aging drops slowly until it eventually levels out to some low value.  For your unit, that could take a few months, maybe more.  They also may not return to the exact same frequency after a power cycle.  This is called 'retrace'.  It's not listed on the datasheet for this model.

Ed
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: OCXO calibration
« Reply #6 on: September 17, 2017, 04:11:47 am »
I have a GPSDO thats on a bare PCB and I am debating how to enclose it. What are peoples feelings on ventilation? Should the box be completely enclosed? It has a Bliley OCXO and I have noticed that when I have my fan going near the GPSDO it seems to go into holdover more frequently. Right now its actually in a cardboard box. cardboard is actually a better insulator than metal when its shut but I likely want a metal box ultimately, when I have figured out exactly what is going to be in there, I may add some DDS functionality..

Im wondering if it should be completely closed. Likely it should be, I am guessing.

The temperature when its operating normally (as recorded by its onboard sensor) is usually between 41 and 50 degrees depending on how its enclosed. The voltage going to the oven changes quite a bit.
« Last Edit: September 17, 2017, 05:35:04 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: OCXO calibration
« Reply #7 on: September 17, 2017, 04:38:03 am »
As texaspyro said, GPS by itself has a stability of about 1e-8, i.e. 10 PPB for a one second measurement with no overall frequency error.  A 10 sec. gate time would give you 1 PPB.

No, the frequency outputs of the Ublox GPS module are not locked to anything.  If they are off 1E-8,  averaging them for however long will still be off 1E-8.   You cannot use the Ublox frequency outputs to meaningfully set an OCXO.

The Ubkix 1PPS output is slaved to GPS.  Averaging that will reduce the uncertainty.  To use the Ublox to set an OCXO, you will need to divide the OCXO down to 1 Hz and compare it to the Ublox 1PPS and deal with the sawtooth error.

The UCT-8663 DOCXO (at the time 10 for $100) that I put in my HP-53132A on one of Gerry Sweeny's boards is drifting less than 1E-10 per month.  Some of that drift is the reference and DAC on the board.   Some of the 8663's from that same lot have drifts in the low E-11 / month range.   It took it around 2 months for it to settle down to that rate.

 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 2260
  • Country: ca
Re: OCXO calibration
« Reply #8 on: September 17, 2017, 06:36:10 am »
As texaspyro said, GPS by itself has a stability of about 1e-8, i.e. 10 PPB for a one second measurement with no overall frequency error.  A 10 sec. gate time would give you 1 PPB.

No, the frequency outputs of the Ublox GPS module are not locked to anything.  If they are off 1E-8,  averaging them for however long will still be off 1E-8.   You cannot use the Ublox frequency outputs to meaningfully set an OCXO.

The Ubkix 1PPS output is slaved to GPS.  Averaging that will reduce the uncertainty.  To use the Ublox to set an OCXO, you will need to divide the OCXO down to 1 Hz and compare it to the Ublox 1PPS and deal with the sawtooth error.

Thanks for the clarification.  I should have mentioned that I was thinking about the 1 PPS.  Are you sure about the frequency error of the other outputs?  I've got a ublox app note "GPS-based Timing Considerations with u-blox 6 GPS receivers" that says that the frequency error of an 8 KHz output is only 6.2e-11.  I'd assume (yeah, I know) that newer timing receivers would be at least that good.  Numbers like that will change from test to test and will eventually average out to zero.  I don't have a lot of experience with ublox boards.

Quote
The UCT-8663 DOCXO (at the time 10 for $100) that I put in my HP-53132A on one of Gerry Sweeny's boards is drifting less than 1E-10 per month.  Some of that drift is the reference and DAC on the board.   Some of the 8663's from that same lot have drifts in the low E-11 / month range.   It took it around 2 months for it to settle down to that rate.

I hate those UCT oscillators!  I've never been able to figure out if they're actually clones of the Oscilloquartz units or not.  The Oscilloquartz 8663 isn't a DOCXO, it's just an OCXO.  Have you found an actual datasheet for the UCT unit?  In any case, your tests show that the UCT-8663 sounds like an upgrade path if Codebird's Isotemp OCXO isn't adequate.

Ed
 

Offline 4CX35000

  • Regular Contributor
  • *
  • Posts: 107
  • Country: gb
Re: OCXO calibration
« Reply #9 on: September 17, 2017, 07:38:12 am »
I've built myself a little precision frequency reference. It uses an OCXO. Specifically, this one:

The purpose of this box is testing cheap, super-low-cost frequency counters off of eBay so that I can either be impressed with their low-cost performance or mock them for including useless extra digits, depending on results. I am curious to see how the El Cheapo bare-PCB counters compare with the thousand-pound-plus professional counters.

Whatever you do it will involve a external reference to do the check which receives on of the standard services such as GPS, MSF, DCF77 (Germany) or Droitwich. There are a number of circuits which are used to receive the Droitwich (Radio 4) reference on the internet, one below is one I came across a while back.

Article which appeared in a magazine is probably a good project to build.
http://www.epemag3.com/lib/free_projects/lab_equipment/0602-%20Frequency%20Standard%20Generator.pdf

If you do build the above circuit then for checking simply use a frequency counter with the 10 MHz from the OCXO on the counter reference input and the 100 KHz from the Droitwich standard on the input. The reading should be 100 KHz exactly and adjust the reference OCXO to the until the counter reads 100 KHz exactly. I suggest you leave the OCXO and Droitwich reference to run for a while for it to fully warm up and then do the check and adjust every so often until your happy with he result.

Afterwards as a future hobby project you can look at using the Droitwich reference as a active reference which automatically adjusts the OCXO as this will drift depending on operating conditions such as room temperature and OXCO oven stability.

As for a counter they tend to be similar. I have a Keithley 776 here and a Thurlby Thandar TF830, and both are reasonable for what I'm doing. But there are plenty of Racal counters on eBay going for around £100.
« Last Edit: September 17, 2017, 09:08:24 pm by 4CX35000 »
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 9410
  • Country: gb
Re: OCXO calibration
« Reply #10 on: September 17, 2017, 10:17:25 am »
As texaspyro said, GPS by itself has a stability of about 1e-8, i.e. 10 PPB for a one second measurement with no overall frequency error.  A 10 sec. gate time would give you 1 PPB.

No, the frequency outputs of the Ublox GPS module are not locked to anything.  If they are off 1E-8,  averaging them for however long will still be off 1E-8.   You cannot use the Ublox frequency outputs to meaningfully set an OCXO.

The Ubkix 1PPS output is slaved to GPS.  Averaging that will reduce the uncertainty.  To use the Ublox to set an OCXO, you will need to divide the OCXO down to 1 Hz and compare it to the Ublox 1PPS and deal with the sawtooth error.

https://www.u-blox.com/sites/default/files/products/documents/Timing_AppNote_%28GPS.G6-X-11007%29.pdf

The OP's calibration requirements aren't particularly demanding.
« Last Edit: September 17, 2017, 10:30:34 am by Gyro »
Best Regards, Chris
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 4137
  • Country: gb
Re: OCXO calibration
« Reply #11 on: September 17, 2017, 10:30:41 am »
Those Trueposition GPSDOs that are popping up on eBay are looking like a good proposition right now if yu need a budget GPSDO, accuracy would seem to be good (are they supported by LadyHeather yet?).

Racal counters, you can pick up bargains, I got two 1998s for ~£120 inc shipping, they needed a 630mA slow blow fuse each and both had the high stability OCXO fitted, last cal dates were around 2007, they've been checked against various 'standards'  (Radio 4 off air, three different GPSDOs, two rubidiums) and, despite being ten years out of cal, still remarkably accurate to the best of my ability to check thought i realise I can't provide calibration crtificates for any of my standards, my logic tells me that if they all came from different sources and all agree to a few tenths of a Hz then they're reliably accurate enough for my needs.

The U-Blox was the least stable of the lot when triggering my coutners, there was a good amount of wander but an interesting snippet of information in the datasheet and the U-Blox software indicates that the can be configured to discipline an off board OCXO if you add a cheap DAC to them and know the specifications of the OCXO
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: OCXO calibration
« Reply #12 on: September 17, 2017, 01:50:24 pm »

I hate those UCT oscillators!  I've never been able to figure out if they're actually clones of the Oscilloquartz units or not.  The Oscilloquartz 8663 isn't a DOCXO, it's just an OCXO.  Have you found an actual datasheet for the UCT unit?  In any case, your tests show that the UCT-8663 sounds like an upgrade path if Codebird's Isotemp OCXO isn't adequate.


I don't think UCT every built their own oscillators.   My guess is they are re-branded Oscilloquartz units.  My UCT rubidium standard has an LPRO-101 in it.

I've put the UCT units in a lot of frequency counters (mostly Tek DC5010's) and have been very surprised at how little they drift.  Also how little they seem to be affected by temperature.
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 3012
  • Country: gb
Re: OCXO calibration
« Reply #13 on: September 17, 2017, 02:46:58 pm »
Quote
The purpose of this box is testing cheap, super-low-cost frequency counters off of eBay so that I can either be impressed with their low-cost performance or mock them for including useless extra digits, depending on results. I am curious to see how the El Cheapo bare-PCB counters compare with the thousand-pound-plus professional counters.

Your OCXO really only has to be stable over time and temperature to test cheapo counters. It doesn't have to be calibrated to be very close to 10MHz. In other words, in your house, your OCXO defines what 10MHz is. Just warm up and align the cheapo counter references to agree with your OCXO and then watch how much they drift the next time you switch them on. Also see how much they drift on a cold day to a hot day or a cold morning to a warm evening. The cheapo counter will probably drift over 10Hz at 10MHz in these tests. Your OCXO should drift over 100 times less than this so I don't see the point of calibrating your OCXO or even buying a GPSDO. It won't add much value to these tests because a cheapo counter will have so much drift anyway.
« Last Edit: September 17, 2017, 02:51:32 pm by G0HZU »
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: OCXO calibration
« Reply #14 on: September 17, 2017, 02:57:53 pm »
Lady Heather supports the Ublox and Venus/Skytraq devices.  The Venus timing receivers have an advantage in that their internal clock is 120 MHz and the sawtooth error is in the +/- 6 ns range.

There can be issues using multiple GNSS systems for precision timing.  The different GNSS systems have slightly different implementations of the second / UTC.   I would avoid enabling multiple GNSS systems for a precision timing application.   I seem to remember Ublox mentioning this.
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: OCXO calibration
« Reply #15 on: September 17, 2017, 03:18:43 pm »
Yes, I've read that too.

 Ublox's defaults to the GPS (USNO) system but lets people pick another one (but not two at the same time) because of the timing difference.

Quote from: texaspyro on Today at 08:57:53
Lady Heather supports the Ublox and Venus/Skytraq devices.  The Venus timing receivers have an advantage in that their internal clock is 120 MHz and the sawtooth error is in the +/- 6 ns range.

There can be issues using multiple GNSS systems for precision timing.  The different GNSS systems have slightly different implementations of the second / UTC.   I would avoid enabling multiple GNSS systems for a precision timing application.   I seem to remember Ublox mentioning this.
« Last Edit: September 17, 2017, 05:38:18 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: OCXO calibration
« Reply #16 on: September 17, 2017, 05:23:33 pm »
Frequency counters should be adjustable so that youre not displaying extra digits when you don't have accuracy to base them on.

Quote from: G0HZU on Today at 08:46:58
Quot
e
« Last Edit: September 17, 2017, 05:40:50 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Offline lars

  • Regular Contributor
  • *
  • Posts: 132
  • Country: se
Re: OCXO calibration
« Reply #17 on: October 06, 2017, 06:08:32 pm »
About GPS module jitter:

The Venus per datasheet for 838-T has an internal (free running) oscillator at about 81 MHz and places the frequency outputs including the PPS on the nearest edge of the 81 MHz. This gives a ripple of +-3ns (6ns p-p). This is also what I have measured, but be aware that it might do some nasty excursions sometimes.

See: https://www.febo.com/pipermail/time-nuts/2016-October/101037.html

The uBlox NEO-6M, NEO-7M and LEA-6T I have tested all have about 21ns p-p jitter. As far as I know both frequency outputs are locked to the GPS but have the 21ns p-p jitter and doesn´t get lower if you use 8MHz instead of 10Mhz out but it looks better on an oscilloscope but you are just cheated.

About the 8663 OCXO:

As far as I known it is a very good double oven XO. In several data sheets for GPS clocks from Oscilloquartz it is mentioned as double oven. My own tests also indicate a double oven from current consumption and temperature stability.


Lars

 
The following users thanked this post: Johnny B Good

Offline lars

  • Regular Contributor
  • *
  • Posts: 132
  • Country: se
Re: OCXO calibration
« Reply #18 on: October 06, 2017, 06:11:19 pm »
For me calibration is about trust. If I measure something I want to trust the result to be as good as I think. To know that, I check (calibrate) my instruments or references against something else that I trust (that is know how accurate it is or said in another way know the measurement uncertainty). If I know someone else has promised how accurate it is and I trust the source I can use that. That is if I buy something from a manufacturer with a specification I can trust it if I trust the manufacturer. It will of course be a small risk my unit is bad for some reason. But this is also true if I calibrate something (by myself or external). Almost all external calibration use a confidence of 95% for that reason. That is, it is 95% chance it meets the uncertainty specification.

For the Isotemp 143 OCXO we have a datasheet from a reliable manufacturer but the problem is the source (used from eBay). Can we trust it?? If I bought it new I could be quite sure it met the spec of slightly better than 1ppm (1uHz/Hz). If that is enough for me I would trust it for further measurements like checking frequency counters. If I need better than the spec I must check it against a better source.

For me the best cheap source is a GPS module as the uBlox NEO6M, LEA-6T or similar and a 10MHz OCXO. As Gyro says in reply #2 you just need to divide down the 10MHz and compare it to the output of the GPS module with an oscilloscope. If you have just the 1PPS I would recommend a digital oscilloscope that you can trigger on the 1PPS. I have done this with just the PPS out from ublox modules and it works well if you divide down the 10MHz at least 10 times. Dividing down 16 times also works well (the output of the divider needs to be an integer of 1Hz).

Using radio or TV signals also worked for me but you need to be sure the transmitter really are locked to eg GPS. I learned this the hard way before I went to only GPS. Another problem I experienced is radio transmitter closing down in a quite short time. For the TV the change from analog to digital made the before long term stable signals almost useless.

After having used GPS to calibrate and adjust my 10MHz references I decided to build a GPSDO as I was tired of adjusting (and I had an Arduino I wanted to learn to program). With a GPSDO I don´t have to adjust or know the error. Of course a GPSDO is not perfect and you need to know the measurement uncertainty even for your GPSDO see for example:

Two links about GPSDO´s as frequency standards and traceable calibration:
http://tf.nist.gov/general/pdf/2289.pdf
http://www.gps.gov/cgsic/meetings/2016/gust.pdf

If you need 1ppb (1E-9 or 0.001ppm) most GPSDO with OCXO´s if showing locked will be ok. But if you need more with good confidence you probably have to understand the specs, if they say enough ? or try to understand the TIC and DAC values to get a feel for the stability.

Note that ADEV is stability and not measurement uncertainty. But as the GPS is locked to a much more accurate reference ADEV can be used to estimate the uncertainty. My own rule of thumb is that the expanded uncertainty (95% confidence) is 2-10 times the worst ADEV at Taus longer than the gate time. To be conservative I normally use 10 times.

Lars
?
 

Offline lars

  • Regular Contributor
  • *
  • Posts: 132
  • Country: se
Re: OCXO calibration
« Reply #19 on: October 06, 2017, 06:17:19 pm »
If you have a GPS module with 1PPS and an 10MHz oscillator with voltage control input I should say it is easy to build a GPSDO for less than 10USD of components for the GPSDO controller.

Here is my simple but well working design that works for all kinds of 10MHz oscillators from XO to rubidiums:

https://www.eevblog.com/forum/projects/lars-diy-gpsdo-with-arduino-and-1ns-resolution-tic/

In the last pages of the instruction you can also see examples of what stability I have achieved with two different oscillators. On page 15 you find a list of useful links to general GPSDO topics, other designs and measurements. On page 8-9 the software including the PI-loop is described.

In hold mode the controller works as an 9 digit per second 10+-0.05 MHz counter and output the delta frequency every second on the serial line, making it easier than using an oscilloscope, with just two HCMOS IC and an Arduino Pro Mini and a few passive components.

Lars
 
The following users thanked this post: edavid

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 2260
  • Country: ca
Re: OCXO calibration
« Reply #20 on: October 06, 2017, 07:06:00 pm »
About the 8663 OCXO:

As far as I known it is a very good double oven XO. In several data sheets for GPS clocks from Oscilloquartz it is mentioned as double oven. My own tests also indicate a double oven from current consumption and temperature stability.
Lars

Can you share any of these documents (or links to them)?  The Oscilloquartz 8663 data sheet doesn't specify single or double oven, but other Oscilloquartz data sheets, e.g. 8660, 8682, and 8863, specify double oven.  Read those model numbers carefully - they're easy to mix up.  From this, I assumed that the 8663 was a single oven design.

Ed
 

Offline lars

  • Regular Contributor
  • *
  • Posts: 132
  • Country: se
 
The following users thanked this post: edavid, kj7e

Offline orin

  • Frequent Contributor
  • **
  • Posts: 445
  • Country: us
Re: OCXO calibration
« Reply #22 on: October 06, 2017, 07:24:42 pm »
Buy yourself a cheap Ublox GPS module off ebay. In brief:

- Divide your OCXO down to 1MHZ.
- Set one of the Ublox module frequency outputs to 1MHz.
- Compare the two signals on a scope, ignoring any jitter, just look at the long term roll between the two traces.
- Tweak the frequency adjustment voltage of the OCXO for zero roll.
- Repeat as often as you feel necessary.



I second the use of a scope to do the comparison.  No need for fancy counters.  In fact, it's the method used by HP in some of their adjustments.
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 2260
  • Country: ca
Re: OCXO calibration
« Reply #23 on: October 06, 2017, 07:43:14 pm »
Here are three links to datasheets from Oscilloquartz describing 8663 as a double oven oscillator:

http://www.sungwhatech.com/product/pdf/08.Time%20&%20Frequency%20Systems/OSA%204500%20OEM%20Clock.pdf
http://www.sungwhatech.com/product/pdf/03.GPS%20Systems/OSA%205200B%20GPS%20Clock.pdf
https://telcogroup.ru/files/pdfs-vend/oscilloquartz/Star-glonass_da_ed03.pdf

Lars

Thanks for the links.

Boy, that's confusing.  What are you supposed to believe?  The datasheet for the oscillator that doesn't say double oven, or the datasheet for other devices that does??  |O

Ed
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: OCXO calibration
« Reply #24 on: October 06, 2017, 08:40:51 pm »
As far as I know both frequency outputs are locked to the GPS

The frequency outputs are not locked to GPS.  They are just divided down from the free running GPS board oscillator via a DDS.   If they could lock the osc and GPS together (like the Thunderbolts do), then there would not be any 1PPS sawtooth error.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf