Author Topic: Rb Frequence Standard or a GPS disciplined for the home lab? Both is better  (Read 5486 times)

0 Members and 1 Guest are viewing this topic.

Offline peluleTopic starter

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
I have followed that thread, weather to use a Rubidium Frequency Standard or a GPS Disciplined one for my home lab?
https://www.eevblog.com/forum/testgear/rubidium-frequency-standard-or-gps-disciplined-ones/msg221974/#msg221974

I had to make the same dicision and decided for the GPS version.
I got a complete VE2ZAZ GPS-Derived 10 MHz Frequency Set for quite low budget including shipment.
A quite good business, even that set is a bit old fashioned already. It including following parts:
  • VE2ZAZ PIC based board
  • Motorola Oncore UT+ GPS Receiver (unfortunatley rev. V3.1, not V3.2, some may know the difference)
  • Motorola Timing 2000 GPS Antenna
  • IsoTemp 10 MHz OCXO134-10
  • 15 m RG58 antenna cable
I build it up and it worked fine.

Next I planned to put it into a not anymore used SCSI External Drive case (fully metal, with FAN and power supply...)

A friend from China heard about my idea and offered me a used DATUM LPRO-101 Rubidium-Oscillator (10 MHz).
He metioned, he doesn't know anything about the condition and may not able to test it.
So, if I would take the risk, he would send it for free, I just need to pay the shipment (not that much).
When it arrived I tested it and found it at a quite good condition.
For that fellows, understanding the specific data, I have them listed here:
   All frequency measured with my most accurate tool, a HP 5316B (calibrated 2 month ago, with option 004, High-stability oven time base), operating for at least >24h.
   All data measured @ 20.0 V:
        Cold: 1.43A, F ~ 10.000150 MHz
        Locked: 0.48A, F ~ 10.000001 MHz
        LAMP Voltage: 7.82V (normal conditions are: operating >3V, healthy >6V … 9V, as higher as better)
Based on DATUM LPRO-101 Service Guide this proofs, the device is in the mid of its lifetime and able to work for many years.
So I had both options now.

To build two frequency standards and two enclosures would double the efforts without an advantage for my (I need just one, but as precise as possible).
The LPRO-101 would need some additional circuits to be useable (power supply, output signal driver...), most of was already available at the VE2ZAZ kit. The IsoTemp 10 MHz OCXO134-10 has less short term stability than the Rb oscillator, the both solution would have advantages and disadvantages.
Searching the WWW for tips to decide I found the great article of Ulrich Bangert regarding the requirements stbility of oscillators and frequency standards for the AMSAT Mars mission(unfortunately in German): http://www.bartelsos.de/index.php?dl_file=77498ICTFZ
Ok, I probably never will fly to the Mars, but to achieve the required precision sound ok for me (also a time-nut?)

So my decision was done: combine the Rb oscillator (its excellent short term stability) with the GPS (to correct the aging of the Rb oscillator / get also long term stability). So I would get the "best out of both worlds" and probably also archive the "best for my money I may get" Frequency Standard.

Just some important questions had to be answered to achieve the best GPS 1pps precision:
- the GPS receiver need to be set into "Position Hold Mode" immediately after power-on
- it has to be ensured, the GPS receiver 1pps signal is never corrupted (in case there is a risk for, it should switches off)
     In fact, if the GPS receiving conditions are not the best, so the stability of the 1pps may become less to the Rubidium osicllator
- I want to have a stand alone and unobserved operating system (I assumed, tracking would need some time/days)
- it should an optional possibility to track/log the data, especially the 1pps signal quality (interface of the GPS)
     Best would be an USB interface.

I had an old NEC (Renesas) "K_Line - Play it! - 2" 78K0 Demonstration kit laying around, offering all I need - 2x12 Character LCD, 2 on-board UARTs (one connected to a FT232 USB/RS232 converter) and several on-chip peripheral (ports, ADC, timers).

Block Diagram:


The Rb-Source has a high short term stability and just slowly drifts due to aging of the RB lamp.

The VE2ZAZ kit saves the latest stable V(tuning) value into its FLASH memory - to ensure the return to last valid tuned frequency. As long as the 1pps signal and the Rb frequency differ too much, the V(tuning) is not changed (but may have a negliable tempCo).
So, even there is no valid (or a not sufficiant) 1pps signal from GPS, the frequency keeps stable and acceptable precise.
The whole VE2ZAZ/GPS  kit just tracks the Rb frequency towards the frequency of the high precision Ceasium Standard of the GPS system (continously calibration).

To ensure, the Rb frequency is not influenced negative by the kit, I have choosen very strict rules to enable the 1pps signal, which is by default OFF (means no tracking).
Before the 1pps signal becomes active (GPS locked - also shown by a green LED at the front) ALL below rules have to be fulfilled:
  • The Rb oscillator is locked (also shown by a yellow LED at the front)l
  • The GPS has detected no error (f.e. no signal at the antenna)
  • The GPS has archieved the "Hold Position", means best possible accuracy of the 1pps signal
  • The GPS Time RAIM algorithm messages no "out of limit alarm", limit is set to 300n)
  • The GPS uses at least 5 satellites (of max. 8 possible) to evaluate the position & timing

I am in the final stage with the MCU software - hope to finish it shortly.
There is just one final question, I am not able to answer yet.
How to measure/verify the precision. I have no such equipment to my access.

Here are some pics of a very similar to mine encloser, I decided to use:
Here a link to a similar enclosure: http://www.interloper.com/products/product-details.php?productid=100863&cat=102

To be continued...

« Last Edit: November 01, 2016, 09:39:24 pm by pelule »
You will learn something new every single day
 
The following users thanked this post: jonovid

Offline peluleTopic starter

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Some pictures.
The Antenna construction is temporary until the design is finished.

To be continued...
You will learn something new every single day
 

Online Vgkid

  • Super Contributor
  • ***
  • Posts: 2727
  • Country: us
Nice job. Do you plan on keeping that anti static bag on the antenna ?  >:D
If you own any North Hills Electronics gear, message me. L&N Fan
 

Offline peluleTopic starter

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
I thought it may boost the antenna. :-DD
The antenna will be mounted on the roof next spring (the best place whis most free air angle and no reflections).
To dangerous now during autumn/winter.
You will learn something new every single day
 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
It's a good thing that you didn't use an antistatic bag made out of metallized plastic.  You wouldn't have much signal left!

Do you need to make any changes to the VE2ZAZ code to match it to the LPRO?  To get best performance, the time constant and maybe damping need to be matched to the oscillator.

Ed
 

Offline peluleTopic starter

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Quote
Do you need to make any changes to the VE2ZAZ code to match it to the LPRO?  To get best performance, the time constant and maybe damping need to be matched to the oscillator.
Didn't need to make any changes at the code/firmware. Just (still) optimize the regulation contol constants:
cycle size = 128 (I plan to increase to 512 as the rubidium orcillator is very stable)
Lock limit = 100 (will be reduce later to 50)
holdover = 20 (fits to the Oncore stability)
You will learn something new every single day
 

Offline peluleTopic starter

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Intermediate update:
System is running non-stop since 3 month now.
Meanwhile I have optimized the VE2ZAZ regulation contol constants:
  cycle size = 1024 (means 16*1024 sec cycle time)
  Lock limit = 5
  holdover = 10
Used it already to adjust the TCXO of my frequency counter (HP5316B with opt 4 & opt 1)
Now just need to finish (drill) the front plate and to mount the atenna on my roof.
For both I need the warmer days as my mechanic tooling is in the garage.

PeLuLe
You will learn something new every single day
 

Offline ap

  • Frequent Contributor
  • **
  • Posts: 286
  • Country: de
    • ab-precision
Does anbody know the drift rates of the LPRO Rubidium standards. It is not specified in the manual.
Reason being, this is calibrated against a GPS receiver, the deviation can be adjusted to below 0.01ppm (<0.1Hz deviation from each other). This is beyond what is needed, and minimizing the effort to calibrate this is one thing, but knowing what it is spec'ed to drift in a year is also important. Also, this unit is not on all the time but only when specific calibration work is done, so the power-on hours are probably more relevant.
Metrology and test gear and other stuff: www.ab-precision.com
 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
The LPRO drift is specified on various datasheets as < 5e-11/month.  Actually, that's the same as many Rb standards.

Ed
 

Offline ap

  • Frequent Contributor
  • **
  • Posts: 286
  • Country: de
    • ab-precision
Thanks, but those I found do not contain this info. can you give a link please or so.
Metrology and test gear and other stuff: www.ab-precision.com
 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
I found them somewhere, but now a quick search doesn't find them anywhere!   :-//  I must have dug them out of the Wayback Machine.

Here they are.

Ed

P.S.  Datum created the LPRO.  Later, Datum was bought up by Symmetricom.
 

Offline ap

  • Frequent Contributor
  • **
  • Posts: 286
  • Country: de
    • ab-precision
Great, thanks, these I didn't find.
Metrology and test gear and other stuff: www.ab-precision.com
 

Offline peluleTopic starter

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
2nd Intermediate update:
GPS Antenna is on the roof - excellent receiving (min 5 sattelites, most the time 7 or 8).
Reworked the 24V DC power supply for the RB Reference - changed it from a linear to a DC/DC Buck Converter (XL4015 module).
System runs now a longterm test.
Next (and last) step is to 3D print a nice front plate.
/PeLuLe
You will learn something new every single day
 

Offline peluleTopic starter

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
3rd intermediate update:
After 1000 hrs non stop operation I change back from the DC/DC converter to the linear power supply.
The received need ~4x cycles to get the GPS looked. Also the lock is lost time to time, when number of visible sattelites is 5 ony.

Print 1st version of a front plate in white PLA. Not all holes fit perfectly (needd some manual correction).
My plan is to print a 2nd version in gray ABS.
/PeLuLe
« Last Edit: March 23, 2019, 07:27:03 pm by pelule »
You will learn something new every single day
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf