Author Topic: TruePosition GPSDO build question  (Read 7099 times)

0 Members and 1 Guest are viewing this topic.

Offline SquiddaddyTopic starter

  • Regular Contributor
  • *
  • Posts: 90
  • Country: us
TruePosition GPSDO build question
« on: January 07, 2019, 12:37:26 am »
I put a TruePosition GPSDO and an Atmega 328 controller in an aluminum case, similar to the BG7TBL units on Ebay.
Those units are not ventilated, so I didn't either, to keep temp fluctuations down.
The room temp is 22-23C or 73F. The GPSDO climbs up to about 38C. At least I haven't been letting it go beyond that. I end up removing the cover.
I'm worried about this running away with heat.
What temperature should I really start worrying, or to keep the frequency stable?
Anyone with a BG7TBL who can tell me how hot they run?

Some pictures of my project.
I have tried with and without the foam cap on the OCXO.
Thanks

 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
Re: TruePosition GPSDO build question
« Reply #1 on: January 07, 2019, 03:37:33 am »
Do NOT put a foam cap over the OCXO.  It's designed to expect a certain amount of heat to escape.  If you prevent that, the unit heats up to a point where the oven throttles back, but the temperature keeps going up!  The oven turns off completely, which means that there's no regulation of the temperature, but the temperature keeps rising just from the normal operating power of the oscillator!  Worst case, the temperature keeps going up until the circuit fries.  Best case is that the unit operates above its design temperature and the performance suffers.  The oven controller could even start oscillating which would really mess up the stability!

Most OCXOs operate their ovens in the 50C to 80C range.  Each unit has its own sweet spot for temperature.  The controller will try to maintain the temperature within a very tight range.  Very high performance units might have a range of 0.001C.   Maximum operating temperature is typically something like 10C below the oven temperature.

Protect the OCXO from drafts (hot or cold), sunbeams, etc., and leave some ventilation room around the outside of your GPSDO, but otherwise, leave it alone.  I see that the NVG47B oscillator is available in three operating temperature values from 50C to 85C.  Unfortunately, there's no easy way to tell what the maximum operating temperature of the NVG47A oscillator in the TruePosition is.  But 40C near the oscillator shouldn't be a problem.  Temperatures over 50C would require some careful investigation.

Ed
 

Offline SquiddaddyTopic starter

  • Regular Contributor
  • *
  • Posts: 90
  • Country: us
Re: TruePosition GPSDO build question
« Reply #2 on: January 07, 2019, 12:54:30 pm »
I'll ditch the foam cap. I thought I read somewhere that a foam cap would help. I guess you can't believe everything you read on the internet.  :)
There is a CTS 196 OCXO on this TruePosition as it's a newer 2011 model.
That temp reading is what's reported from the board in the datastream. I'll put a temp sensor inside the case to get the surrounding air temp, and monitor for a while.
These things are pretty slick for the price. After I did a 12 hour survey to get position, it has stayed locked for several days, usually getting 5 to 9 satellites. 5 or 6 at night and 8 or 9 during the day.
Took me a while to get the software written the way I wanted it, but it worked out pretty good.

I appreciate your advice Ed!
 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
Re: TruePosition GPSDO build question
« Reply #3 on: January 07, 2019, 05:25:08 pm »
I'll ditch the foam cap. I thought I read somewhere that a foam cap would help. I guess you can't believe everything you read on the internet.  :)
There is a CTS 196 OCXO on this TruePosition as it's a newer 2011 model.
That temp reading is what's reported from the board in the datastream. I'll put a temp sensor inside the case to get the surrounding air temp, and monitor for a while.

When I got started in this stuff I thought a foam cap would be a good idea.  But when I started reading the Time-Nuts mailing list I saw a message that explained why it was a bad idea.  If you were running an OCXO in a cold environment, the foam cap might be a good idea, but not in a normal temperature environment.

Quote
These things are pretty slick for the price. After I did a 12 hour survey to get position, it has stayed locked for several days, usually getting 5 to 9 satellites. 5 or 6 at night and 8 or 9 during the day.
Took me a while to get the software written the way I wanted it, but it worked out pretty good.

A properly installed and operating GPSDO should stay locked forever.  Any loss of lock should be investigated.  Of course, if your antenna location and satellite visibility are crappy you might have issues.

Quote
I appreciate your advice Ed!

Glad I could help.

Ed
 

Offline jpb

  • Super Contributor
  • ***
  • Posts: 1771
  • Country: gb
Re: TruePosition GPSDO build question
« Reply #4 on: January 07, 2019, 06:21:29 pm »
By way of comparison, my Star4 GPSDO reports (via Lady Heather) its temperature at around 48 to 49 degrees C which I was quite happy with though it is getting a little high - OCXOs generally are specced to either 0 to 50 I think or -25 to 75, that sort of range.

If an OCXO is put into a smallish metal box then it will get quite warm.

I think the crystal itself (within the oven) is kept at around 70 degrees C (though obviously higher for OCXOs specced to go to 75 externally) - it is what ever temperature presents an optimal turning or inflection point.
« Last Edit: January 07, 2019, 06:23:03 pm by jpb »
 

Offline rbm

  • Regular Contributor
  • *
  • Posts: 230
  • Country: ca
Re: TruePosition GPSDO build question
« Reply #5 on: January 08, 2019, 02:15:23 am »
Suiddaddy, nice build!  I'm also in the midst of construction of a Trueposition GPSDO.  I found a perfect enclosure for my TP board, power supply and CPU.  I'm basing my CPU drive off of pigrew's STM32 Blue Pill based design.  Since you are using an ATmega based CPU, I assume you are using Arduino based code.  Is the code you are using based on one of the existing Arduino designs?  Or is it your own design?
- Robert
 

Offline SquiddaddyTopic starter

  • Regular Contributor
  • *
  • Posts: 90
  • Country: us
Re: TruePosition GPSDO build question
« Reply #6 on: January 08, 2019, 03:38:39 am »
It's my own code. I used some ideas from Pigrew's code to parse the packets, but i did the rest myself. My Trueposition won't communicate on the ttl port because i had a faulty usb/ttl adapter, and it damaged it.
So i had to incorporate a Max232 serial chip and set up 2 serial ports. And i used an I2C lcd display.
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: TruePosition GPSDO build question
« Reply #7 on: January 08, 2019, 04:08:14 am »
Do NOT put a foam cap over the OCXO.  It's designed to expect a certain amount of heat to escape.  If you prevent that, the unit heats up to a point where the oven throttles back, but the temperature keeps going up!

That's not necessarily true.  I have several GPSDOs where the manufacturer insulated the OCXO.  Most used a thin flexible foam (often open at the top) but a couple have much more substantial insulation.

But I would not recommend insulating a random OCXO...
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: TruePosition GPSDO build question
« Reply #8 on: January 08, 2019, 04:28:11 am »
I tried that too, but it didn't seem to be helpful. I was expecting the temperature to remain the same but it went up higher. So the TP case I'm working on is ventilated and has a power supply. I'm just going to try to moderate the worst external extremes from getting to it.
"What the large print giveth, the small print taketh away."
 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
Re: TruePosition GPSDO build question
« Reply #9 on: January 08, 2019, 05:20:35 am »
Do NOT put a foam cap over the OCXO.  It's designed to expect a certain amount of heat to escape.  If you prevent that, the unit heats up to a point where the oven throttles back, but the temperature keeps going up!

That's not necessarily true.  I have several GPSDOs where the manufacturer insulated the OCXO.  Most used a thin flexible foam (often open at the top) but a couple have much more substantial insulation.

But I would not recommend insulating a random OCXO...

I saw one case of the "open at the top" external insulation and wondered why it was there.  Maybe it was to help the oven heat up to meet the low temperature limit for the unit.  In any case, it's one thing for the manufacturer to do it, but something quite different for us to do it.

Ed
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: TruePosition GPSDO build question
« Reply #10 on: January 08, 2019, 05:53:10 am »
So the TP case I'm working on is ventilated and has a power supply. I'm just going to try to moderate the worst external extremes from getting to it.

You could try using Lady Heather's temperature control PID to PWM a fan and keep the temperature constant...
 

Offline SquiddaddyTopic starter

  • Regular Contributor
  • *
  • Posts: 90
  • Country: us
Re: TruePosition GPSDO build question
« Reply #11 on: January 08, 2019, 12:46:09 pm »
I watched the YouTube video of CircuitValley's build, and he crammed so much into his little enclosure that he was running in the 50C+ range. I run no where near this, usually 35-38C. 40C when the sun warms the room up late in the day. He'll probably cook those boards in no time.
Part of my heat is the 7805 regulator in there to power the AtMega328, the Max232, and the LCD. It's using a clip on heatsink, so I may bolt it to the enclosure.
I have the 2011 board with 2 10MHz outputs, and a G revision of the Furuno GPS, which works out well, as I can feed the counter and the signal gen. No need for distribution yet.

 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: TruePosition GPSDO build question
« Reply #12 on: January 08, 2019, 10:19:53 pm »
You wouldn't have any photos representative of what they do, do you?


The one photo I have of the interior of the True Position LM-300 unit (from ebay) that actually shows the GPSDO part and its context) shows it right next to an (exhaust?) fan.

Do NOT put a foam cap over the OCXO.  It's designed to expect a certain amount of heat to escape.  If you prevent that, the unit heats up to a point where the oven throttles back, but the temperature keeps going up!

That's not necessarily true.  I have several GPSDOs where the manufacturer insulated the OCXO.  Most used a thin flexible foam (often open at the top) but a couple have much more substantial insulation.

But I would not recommend insulating a random OCXO...
« Last Edit: January 08, 2019, 10:33:39 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Offline SquiddaddyTopic starter

  • Regular Contributor
  • *
  • Posts: 90
  • Country: us
Re: TruePosition GPSDO build question
« Reply #13 on: January 10, 2019, 04:15:07 pm »
Something else I noticed reading all of the TruePosition discussion on here is a command to set the cable delay of the antenna.
I have 50ft of cable going to the outside timing antenna. Would it matter if the delay was added, or is it small enough to not bother? Does the cable delay affect the 10MHz output, or just the 1PPS signal? I'm not planning to use this as any time source, just a frequency reference.
I don't know how to get the delay figure anyway.

Also, the DAC value to the OCXO on it still changes. Once this thing locks to GPS, shouldn't the DAC stay put for long periods of time? Or is it always adjusting? These changes are in microvolts, so I believe it's normal. Currently at 2.071xxx volts, with the last 3 digits changing slightly every 15 seconds or so. And that's with the Atmega328 rounding it to 6 digits. Usually seeing 8 satellites in the sky.
So far, I'm impressed with this thing.
Been powered up a little over a week now. Unfortunately, I have to shut it down to put it on a UPS at some point.  :(
 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
Re: TruePosition GPSDO build question
« Reply #14 on: January 10, 2019, 04:56:50 pm »
You wouldn't have any photos representative of what they do, do you?


The one photo I have of the interior of the True Position LM-300 unit (from ebay) that actually shows the GPSDO part and its context) shows it right next to an (exhaust?) fan.

Do NOT put a foam cap over the OCXO.  It's designed to expect a certain amount of heat to escape.  If you prevent that, the unit heats up to a point where the oven throttles back, but the temperature keeps going up!

That's not necessarily true.  I have several GPSDOs where the manufacturer insulated the OCXO.  Most used a thin flexible foam (often open at the top) but a couple have much more substantial insulation.

But I would not recommend insulating a random OCXO...

Three points to remember:
Under normal operating conditions, this equipment probably operates in an air-conditioned computer room.  Once the unit's temperature stabilizes, it will never vary.  So it doesn't matter if the OCXO is near a fan.

We don't really know what the design criteria for this equipment was - accurate time of day, low phase noise, or rock stable frequency.

The OCXO is operating inside a PLL so frequency changes will be slowly corrected.

As amateurs trying to reuse this stuff, the best we can do is to try and keep the operating conditions as stable as we can.

Ed
 

Online edpalmer42

  • Super Contributor
  • ***
  • Posts: 2351
  • Country: ca
Re: TruePosition GPSDO build question
« Reply #15 on: January 10, 2019, 05:08:21 pm »
Something else I noticed reading all of the TruePosition discussion on here is a command to set the cable delay of the antenna.
I have 50ft of cable going to the outside timing antenna. Would it matter if the delay was added, or is it small enough to not bother? Does the cable delay affect the 10MHz output, or just the 1PPS signal? I'm not planning to use this as any time source, just a frequency reference.
I don't know how to get the delay figure anyway.

Typically, the antenna delay is used if you need your 1 PPS to be synchronized with UTC.  Since there's delay in the antenna cable, you have to compensate for it.  We don't care about such things, so it's almost never used.

Quote
Also, the DAC value to the OCXO on it still changes. Once this thing locks to GPS, shouldn't the DAC stay put for long periods of time? Or is it always adjusting? These changes are in microvolts, so I believe it's normal. Currently at 2.071xxx volts, with the last 3 digits changing slightly every 15 seconds or so. And that's with the Atmega328 rounding it to 6 digits. Usually seeing 8 satellites in the sky.
So far, I'm impressed with this thing.
Been powered up a little over a week now. Unfortunately, I have to shut it down to put it on a UPS at some point.  :(

The OCXO will probably take weeks, if not longer, of continuous operating before it really settles down.  Aging will be high, there may be frequency and/or phase jumps.  During this period, the PLL circuitry will really be working hard and, as a result, the DAC value will bounce around.  How stable the DAC value eventually becomes will depend on the characteristics of your particular OCXO and the output of your GPS receiver.

Ed
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: TruePosition GPSDO build question
« Reply #16 on: January 11, 2019, 02:42:28 am »
You need to know the velocity factor and length of your antenna's coax cable, it could be gotten either from the manufacturer - (or a chart like one at http://rfcafe.com/references/electrical/coax_chart.htm  ) if you know exactly what coax was used- (should be written on it) or it can also be determined experimentally using an oscilloscope



« Last Edit: January 11, 2019, 02:49:15 am by cdev »
"What the large print giveth, the small print taketh away."
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: TruePosition GPSDO build question
« Reply #17 on: January 11, 2019, 02:48:56 am »
You need to know the velocity factor and length of your antenna's coax cable, it could be gotten either from the manufacturer - if you know exactly what coax was used, or determined experimentally.

Lady Heather's cable delay setting command can set the value directly in nanoseconds, or you can specify the value in feet/meters of cable with an optional velocity factor (default Vf is 0.66).  Also your antenna cable delay value should include the delay in the 1PPS wiring...  which probably has a different velocity factor.

One gotcha in setting cable delay values is the sign of the required value.  Some devices use positive numbers, others use negative numbers... and very few manufacturers document what their device expects.
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: TruePosition GPSDO build question
« Reply #18 on: January 11, 2019, 02:55:19 am »
Negative VF?

That would really be an amazing feat wouldn't it be, if a signal could be made to emerge from a transmission line before it went in!

"Deja-vu"

« Last Edit: January 11, 2019, 03:06:17 am by cdev »
"What the large print giveth, the small print taketh away."
 

Offline SquiddaddyTopic starter

  • Regular Contributor
  • *
  • Posts: 90
  • Country: us
Re: TruePosition GPSDO build question
« Reply #19 on: January 11, 2019, 12:39:38 pm »
When I watch the clock output of the TruePosition and the computer, they both switch to the next minute as accurately as my eyes can determine. I'm not going to sweat over nanoseconds.  :-DD
I think I'll leave the settings alone.
Good info on determining the velocity factor though.  :)

Question for @texaspyro, how does LadyHeather know when to drop a satellite off of the list? Does the TruePosition tell you anything in the $SAT messages from the stream? I was going to include satellite data in an array for display on my AtMega328 LCD, but didn't bother because I didn't know how/when to remove them. And I didn't want to overcomplicate it since it's only software for me.
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 4136
  • Country: gb
Re: TruePosition GPSDO build question
« Reply #20 on: January 11, 2019, 01:58:09 pm »
Negative VF?

That would really be an amazing feat wouldn't it be, if a signal could be made to emerge from a transmission line before it went in!

"Deja-vu"

I seem to remember it was a cable problem that caused time travelling neutrinos to be detected a few years ago, maybe they inadvertently messed up the specifications and bought a reel of cable with negative velocity factor ;)

https://www.livescience.com/16207-faster-light-discovery-time-travel.html
 
The following users thanked this post: cdev

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: TruePosition GPSDO build question
« Reply #21 on: January 11, 2019, 02:59:54 pm »
I would like some of that magical cable, please!

"One way to think of this is like a mirror, Lloyd said. You send a message to the mirror, and it reflects it back, but so quickly that "past you" is the one who receives it."

« Last Edit: January 11, 2019, 03:14:50 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: TruePosition GPSDO build question
« Reply #22 on: January 11, 2019, 03:29:19 pm »
It always takes longer for the signal to go through the cable than light  :-\

(1/n)

Even if its a superconductor.  :palm:
"What the large print giveth, the small print taketh away."
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: TruePosition GPSDO build question
« Reply #23 on: January 12, 2019, 01:52:51 am »

Question for @texaspyro, how does LadyHeather know when to drop a satellite off of the list? Does the TruePosition tell you anything in the $SAT messages from the stream?


Heather drops any sats with a sig level of 0.   
 

Offline krulls

  • Contributor
  • Posts: 14
  • Country: us
Re: TruePosition GPSDO build question
« Reply #24 on: May 14, 2020, 12:45:10 am »
Hi - have you published your project details publicly? Reason I ask is I have a True Position board and I tried running the Arduino sketch provided by the Packrats GPS Project people (https://www.packratvhf.com/attachments/article/160/A-Packrat-GPS-Receiver-Project.pdf) and I get nothing but a line of 5x8 pixel blocks on the LCD screen. I haven't received any replies to several emails to the project creators so I'm looking for an alternate solution.
Many thanks,
Steve
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf