Economical option for precision frequency reference?

[edpalmer42]

By the way, Ed - the Lucent seems to be almost an order of magnitude more stable than this thing so far. It's interesting to watch the Lucent OCXO age. The EFC voltage has been following a general upward trend since I powered it on. I guess at some point, this will level off, and at that point the OCXO will be considered broken in?

Not really.  OCXOs age - period.  That's why you need a GPSDO - to lock the OCXO to a stable reference and keep the frequency constant.  What will happen is that the rate of aging will slow down to the specified value for the OCXO or lower.  If you had 100 OCXOs of the same type you'd find that aging could be high or low, positive or negative depending on the individual unit.  Aging could be positive one day and negative the next, even after you've accounted for changes in temperature, supply voltage, humidity (yes!), air pressure (yes!) etc.  Whether the levels are significant depends on the user and the application.

Ed

[motocoder]

By the way, Ed - the Lucent seems to be almost an order of magnitude more stable than this thing so far. It's interesting to watch the Lucent OCXO age. The EFC voltage has been following a general upward trend since I powered it on. I guess at some point, this will level off, and at that point the OCXO will be considered broken in?

Not really.  OCXOs age - period.  That's why you need a GPSDO - to lock the OCXO to a stable reference and keep the frequency constant.  What will happen is that the rate of aging will slow down to the specified value for the OCXO or lower.  If you had 100 OCXOs of the same type you'd find that aging could be high or low, positive or negative depending on the individual unit.  Aging could be positive one day and negative the next, even after you've accounted for changes in temperature, supply voltage, humidity (yes!), air pressure (yes!) etc.  Whether the levels are significant depends on the user and the application.

Ed

I see. Well, the PPS excursions reported by Z38XX have definitely settle down since initial power-on. Now it is maybe  50% of the time within +/-10ns, 70% of the time within +/-20 ns, and it does not ever seem to go above about 60ns, whereas when it was first powered on, it was bouncing between +/- 90ns pretty much constantly.

[Vgkid]

For those who have the rftg-ii-rb, what is the power input pin.

[cncjerry]

I bought the Lucent set and it arrived yesterday.  I added a Garmin external antenna and powered the units using a laptop 19.5v brick until I build a better supply.  They came up in about 15 minutes and after locating the antenna optimally I am getting 6 and sometimes 7 satellites.  I thought they only tracked 5 at a time? Satstat works well and the unit finished the survey.  Tfom  and ffom are as low as they go.  Overall very pleased with how smoothly it all went.

I bought them to sync my counters and spectrum analyzer.  The killer is they were hardly off much at all after all that. 

Jerry

[motocoder]

I bought the Lucent set and it arrived yesterday.  I added a Garmin external antenna and powered the units using a laptop 19.5v brick until I build a better supply.  They came up in about 15 minutes and after locating the antenna optimally I am getting 6 and sometimes 7 satellites.  I thought they only tracked 5 at a time? Satstat works well and the unit finished the survey.  Tfom  and ffom are as low as they go.  Overall very pleased with how smoothly it all went.

I bought them to sync my counters and spectrum analyzer.  The killer is they were hardly off much at all after all that. 

Jerry

They can sync up to 8 satellites at a time.

[motocoder]

I bought the Lucent set and it arrived yesterday.  I added a Garmin external antenna and powered the units using a laptop 19.5v brick until I build a better supply.  They came up in about 15 minutes and after locating the antenna optimally I am getting 6 and sometimes 7 satellites.  I thought they only tracked 5 at a time? Satstat works well and the unit finished the survey.  Tfom  and ffom are as low as they go.  Overall very pleased with how smoothly it all went.

I bought them to sync my counters and spectrum analyzer.  The killer is they were hardly off much at all after all that. 

Jerry

I am very happy with mine as well. It is more than accurate enough to calibrate my frequency counter.

[smgvbest]

For those who have the rftg-ii-rb, what is the power input pin.

Pin 1 is +24VDC and Pin 2 is GND

[smgvbest]

Has anyone used the RTFG.exe on WinXP to configure the RFTGm-II-Rb/RFTGm-II-XO pair?
From the look of it, it was written for Win95 or Windows 3.11.   I can load up an old PC with Win95 if I have to but if anyones gotten it to work on WinXP that would be great to know.

Just mainly looking to program in my cable length or atleast read out what it is.   I've used a FTDI based RS422 interface and can see binary data coming across in teraterm but the RTFG program doesn't seem to connect

I do know using the output of the XO GPD Locked and used on my HP53131A's external ref and then measuring the Rb output I have 10Mhz +/- 200ns

[Vgkid]

I got it powered, using the db15 connector.
Unless i've messed up, do not try connecting the external comms connectors(rs422) to the db9 on the data/power connector unit.
sparks will fly.

[cncjerry]

Today around noon I tripped the breaker powering my ref-0/ref-1 pair.  Prior to this happening my ref-0 was active and transmitting the 15Mhz signal. Ref-1 was in standby. Upon powering the units back up, ref-0 is showing a fault (after the blinking light test) and now ref-1 is active. Ref-0 is remaining in fault state.  Ref-1 is running the survey again.

I looked in the error log and around the time of the failure a "15Mhz output failure" was logged.  Stupid me I cleared the log so I can't read the correct error text.  So I guess what happened is that the ref-0 (which was active) saw the power failure first?  transferred to ref-1?  and now ref-1 will be active until the next failure?  I power cycled them right after the failure so they came up at the same time.  I would have thought ref-0 would have taken over again.  And if ref-1 is active I would have thought ref-0 would be in standby, not fault, no?

Is there a way thru Satstat to have them switch back or is there another way to test the fail-over?  Also, Since the 10Mhz output is on ref-0, does that remain sync'ed when ref-1 is active?

Lastly, I see there is a power-on setting to either have it rerun the survey or use the last known position.  I don't see a way to set it.  Querying the device shows it defaults to rerun.

thanks

[motocoder]

Today around noon I tripped the breaker powering my ref-0/ref-1 pair.  Prior to this happening my ref-0 was active and transmitting the 15Mhz signal. Ref-1 was in standby. Upon powering the units back up, ref-0 is showing a fault (after the blinking light test) and now ref-1 is active. Ref-0 is remaining in fault state.  Ref-1 is running the survey again.

I looked in the error log and around the time of the failure a "15Mhz output failure" was logged.  Stupid me I cleared the log so I can't read the correct error text.  So I guess what happened is that the ref-0 (which was active) saw the power failure first?  transferred to ref-1?  and now ref-1 will be active until the next failure?  I power cycled them right after the failure so they came up at the same time.  I would have thought ref-0 would have taken over again.  And if ref-1 is active I would have thought ref-0 would be in standby, not fault, no?

Is there a way thru Satstat to have them switch back or is there another way to test the fail-over?  Also, Since the 10Mhz output is on ref-0, does that remain sync'ed when ref-1 is active?

Lastly, I see there is a power-on setting to either have it rerun the survey or use the last known position.  I don't see a way to set it.  Querying the device shows it defaults to rerun.

thanks

I don't think any special command is required for it to switch back. For sure it should go back to REF-0 after power cycle. Double-check your interface cable (J5) and really make sure it's screwed in tight. I saw several reports of people having issues with that cable caused by some of the pins being shorter than normal.

[cncjerry]

Wrt my lengthy post above, the fault light on ref-0 never went out.  Ref-1 worked as expected and ref-0 continued to output 10Mhz but not the 15Mhz which continued to be supplied by ref-1.  The 'on' light was lit on ref-1;  ref-0 posted a fault light.

For the heck of it, I pulled the power on ref-1.  The 'no gps' light was lit on ref-0 which continued to display 'fault'.  I would have expected ref-0 to go into hold over and output the 15Mhz, but it didn't.  It seems like it outputs 10Mhz all the time. 

I then power cycled both units. After going thru the blink test, the 'on' light on ref-0 lit and it started to output 15Mhz.  I would have expected it to do this when I pulled the power on ref-1 forcing a fail over, but it didn't.  Also, I had changed the power on default to use the last position so it wasn't doing a survey again.

After about 10 minutes of ref-0 outputting 15Mhz with the 'on' and 'fault' lights lit (prior to full gps lock), ref-0 posted errors in the log:
15Mhz Failure Set
Failed Mode Entered
At that point, ref-1 lit its 'on' light and started outputting 15Mhz.  So I guess ref-0 saw a failure and it failed over to ref-1.

I then cycled the power again.  The errors are still in the log, but ref-0 is now doing a survey and acting OK.  Both units have the 'no gps' and 'fault' lights lit (prior to the survey which is normal) with ref-0 also having the 'on' light lit and outputting 15Mhz.

So if you can follow the confusing sequence above, the questions are:

1) If you pull the power on the active unit ('on' light lit), shouldn't it transfer to the other?  If the unit you pull power from is ref-1 with the gps, it should go into hold-over;  if you pull ref-0 (without gps) shouldn't ref-1 just take over and not go into hold over?  You would lose the 10Mhz from ref-0, of course.

2) Based on the sequence above, I suspect my ref-0 which is posting the '15Mhz failure set' errors to have a problem.  It transferred twice to the ref-1 unit.  At first I thought it was because of the power failure but now I suspect not.

3) Does anybody know how it monitors the 15Mhz output to be able to tell when it has a failure?

4) It seems like the 10Mhz is always running.  Is this gps locked?  If so, I would not expect to see it if the gps wasn't up.

5) Under normal operation, what lights are lit on your units?  Which of the two is the default for 15Mhz?   I thought when my set were working correctly, ref-1 was in standby and ref-0 had the 'on' light lit and was outputting both 10Mhz and 15Mhz.

I am using 19.6V to drive both of them.  I had seen posted that the DC-DC converter takes anything from 18 to like 36V.  I plan to build a better power supply but I am starting to think maybe I am too close to the 18V.

Sorry for the long confusing posts.

Jerry

[Vgkid]

I had the same issue earlier today in which the xo stayed offline, while the rb remained online. Similiar to your a bad piwer connection caused the xo to lose power while the. Rb stayed online. A power cycle fixed that.

[motocoder]

I am using 19.6V to drive both of them.  I had seen posted that the DC-DC converter takes anything from 18 to like 36V.  I plan to build a better power supply but I am starting to think maybe I am too close to the 18V.

19.6V is fine, but the lower the input voltage, the more amperage is required. There are some LEDs near rear left of the unit (looking down from above), just to the right of the DC-DC converter. If power is OK, you should see one green LED blinking at a rate of about once per second. If you see anything else, there is a power fault. Can you check this, especially during start-up sequencing (it draws more power during that phase)?

[cncjerry]

Moto, the internal power lights are fine as you stated, slowly blinking green.

I came down after dinner where before I had recycled the power.  Ref-0 (where I had the fault) was now ok, no fault, and 'on' putting out 15Mhz;  ref-1 was in 'standby' mode.  So it looks like a couple of power cycles plus some  waiting time fixed it.

So not being able to leave well enough alone, I switched the front panel Output level from 17 - 23 on ref-0.  The 'on' lights started blinking in both units and they logged an error that the setting were different.  Ref-0 then failed over to ref-1.  So I set them both back to the same 17 level setting.  I then recycled the units and ref-0, the one I've been having trouble with, is now showing fault again and ref-1 is in 'on' mode.  It posted another 15Mhz failure.

I'm wondering if you have to have the 15Mhz unloaded when it boots?  I'll attempt that next as I try to clear the fault on ref-0.  It might just clear itself.  It shows that it is GPS locked which usually is what clears the fault.

I don't know if I am adding to the wealth of knowledge on these boxes or confusing the subject...

[motocoder]

Moto, the internal power lights are fine as you stated, slowly blinking green.

I came down after dinner where before I had recycled the power.  Ref-0 (where I had the fault) was now ok, no fault, and 'on' putting out 15Mhz;  ref-1 was in 'standby' mode.  So it looks like a couple of power cycles plus some  waiting time fixed it.

So not being able to leave well enough alone, I switched the front panel Output level from 17 - 23 on ref-0.  The 'on' lights started blinking in both units and they logged an error that the setting were different.  Ref-0 then failed over to ref-1.  So I set them both back to the same 17 level setting.  I then recycled the units and ref-0, the one I've been having trouble with, is now showing fault again and ref-1 is in 'on' mode.  It posted another 15Mhz failure.

I'm wondering if you have to have the 15Mhz unloaded when it boots?  I'll attempt that next as I try to clear the fault on ref-0.  It might just clear itself.  It shows that it is GPS locked which usually is what clears the fault.

I don't know if I am adding to the wealth of knowledge on these boxes or confusing the subject...

The output level alarm is well-known. One of the pins on J5 indicates the level setting. How long have you waited to see if the fault clears? I'd wait at least 10 minutes to see if things return to a normal state.

[cncjerry]

I power cycled them again and all is back to normal.  If I had to point to anything I would think the problem is inconsistent application of power.  We'll see if the 15Mhz failure comes back.

Thanks for the pointers.

Jerry

btw:  when ref-0 faulted with the 15Mhz error, I waited about an hour for it to clear without luck.   The good news was it failed-over.  I'd still like to know what the 15Mhz error means and how they track it. 

[cncjerry]

Once the survey is complete and the units have been running for a while, how many sats does it have to lock to maintain its best sync?  I can get 4 sat reliably with my antenna outside my shop;  If put it on my railing three stories up I can get 8.  Just wondering if 4 sats are enough for long term operation.

thanks

Jerry

[Dragon88]

Jerry,

I'll try to answer some of your questions. I've been working on these units quite a bit the past couple weeks, with the goal of getting the two boxes (REF-0 and REF-1) to operate independently of each other. I have been successful in this, and have learned a lot about them in the process.

- Each unit knows the status of the other unit using pins on the interface connector. There is no sophisticated signaling. The logic level of three pins in particular tell the REF-0 to initially turn its outputs on. Either by command or through a fault, the REF-1 can take over and turn its own outputs on. I do not know what your own fault is, but I would recommend removing and re-seating the interface cable several times. There may be corrosion on one of the pins which is causing an intermittent connection.

- The 10MHz output on the REF-0 is a test point, and is not a primary output. It is disciplined in that it is generated from the 5MHz OCXO in the unit, which is disciplined to GPS. However, it's status (on or off) is not controlled like the 15MHz primary output is. It's actually buffered by the same chip that generates the clock signal for the processor. A "test point", nothing more. The signal is quite "dirty" compared to the 15MHz output.

- The REF-0 send no commands at all to the GPS. In fact the serial Tx pin from the REF-0 isn't even connected. The REF-0 monitors the GPS strings (and PPS signal) sent from the Oncore GPS to the REF-1, and looks for key fields to know that the GPS lock is good. If it detects that GPS status is degraded, it will enter holdover mode, and recover on its own when the lock is good again. The REF-1 does not guide this process. All the two boxes exchange between each other is "I'm ok and my outputs are on" or "I'm not ok you turn your outputs on". Again, this is merely done with pins on the interface connector for reliability.

- You should always achieve the best antenna position possible.

If you need any more help with your unit, please PM me. 

Dan

[SoundTech-LG]

That's great.

So if you have the information about "which" pins, and a truth table for them, it would be helpful to many people to post the information here.

Thanks!

[TheSteve]

Bought an NTBW50AA from the 'bay the other day, looking forward to firing it up when it arrives. I have read there is a divide by two on the 9.8304 MHz which I plan to remove if possible to output 19.6608 MHz which my Agilent spectrum analyzer can use. It can also use the even second output which should be fun to try. Seemed to be the cheapest option overall that I could find.

[artag]

I'll try to answer some of your questions. I've been working on these units quite a bit the past couple weeks, with the goal of getting the two boxes (REF-0 and REF-1) to operate independently of each other. I have been successful in this, and have learned a lot about them in the process.

Thank you for publishing this work (and thanks to the other time-nuts people who have provided information). I wanted to get one of these going but the low price is rather spoilt by the postage to the UK - about 3 times the price of the units iirc !

The more you ship the cheaper it is .. though bulk does tend to make sure the customs and excise people notice it. Despite that,  I've bought several to make it worthwhile and am just about to take delivery - so if you're in the UK or nearby and want to share in the oscillatory goodness at a CMOT price, get in touch.
 

[TheSteve]

I recently needed to buy a GPS antenna for my gpsdo and after touring ebay found item 181420921843.
I received it yesterday and can highly recommend it for anyone looking to get a very nice outdoor GPS antenna for a good price.
The listing isn't too great but what you are really getting is a new waterproof Trimble 5 volt 26 db gain antenna, part # 28367-70 and a 5 meter RG316 cable made by Huber Suhner with a TNC on one end for the antenna and an SMA on the other. The cable is very highly quality - no cheap chinese parts.
Pretty tough to beat this combo for 15.99 USD shipped for those in the US.

I have no affiliation with the seller, just like a good deal and this kit is ideal for many gpsdo owners.

edit - aww, they just put the price back to 19.99 for them, still a good deal but not as good.

[motocoder]

I recently needed to buy a GPS antenna for my gpsdo and after touring ebay found item 181420921843.
I received it yesterday and can highly recommend it for anyone looking to get a very nice outdoor GPS antenna for a good price.
The listing isn't too great but what you are really getting is a new waterproof Trimble 5 volt 26 db gain antenna, part # 28367-70 and a 5 meter RG316 cable made by Huber Suhner with a TNC on one end for the antenna and an SMA on the other. The cable is very highly quality - no cheap chinese parts.
Pretty tough to beat this combo for 15.99 USD shipped for those in the US.

I have no affiliation with the seller, just like a good deal and this kit is ideal for many gpsdo owners.

edit - aww, they just put the price back to 19.99 for them, still a good deal but not as good.

The antenna is waterproof, but what about the connector? On the Symmetricon antenna that I have, the connection is made inside a tube, the top of which is sealed with a gasket. So unless you mount it upside down, water can't get in. As you say the listing isn't too good, but based on the picture that's there it looks like water could get into the connector.

[TheSteve]

I recently needed to buy a GPS antenna for my gpsdo and after touring ebay found item 181420921843.
I received it yesterday and can highly recommend it for anyone looking to get a very nice outdoor GPS antenna for a good price.
The listing isn't too great but what you are really getting is a new waterproof Trimble 5 volt 26 db gain antenna, part # 28367-70 and a 5 meter RG316 cable made by Huber Suhner with a TNC on one end for the antenna and an SMA on the other. The cable is very highly quality - no cheap chinese parts.
Pretty tough to beat this combo for 15.99 USD shipped for those in the US.

I have no affiliation with the seller, just like a good deal and this kit is ideal for many gpsdo owners.

edit - aww, they just put the price back to 19.99 for them, still a good deal but not as good.

The antenna is waterproof, but what about the connector? On the Symmetricon antenna that I have, the connection is made inside a tube, the top of which is sealed with a gasket. So unless you mount it upside down, water can't get in. As you say the listing isn't too good, but based on the picture that's there it looks like water could get into the connector.

I'd recommend any outdoor antenna connection be properly sealed in vulcanizing/Self-amalgamating tape.