What GPSDO do you recommend?

[FriedMule]

I have been bitten by high-ish accuracy and I would like to hear what you use or recommend getting for best long time and short time measuring accuracy.
Right now am I playing with a HP Z3805A and a Trimble 10MHz ThunderBolt.

[Conrad Hoffman]

Don't know where it falls in the "great scheme" but I like the Leo Bodnar units because you can program the output frequency over a huge range.

[donlisms]

Bodner.  There is the lesser model and the greater model.  Either one.

[miro123]

Another happy owner of Leo Bodna Precision GPS Reference Clock

[djsb]

+1 on Bodnar.

[FriedMule]

I'll take a look at the Leo Bodnar but what are the advantage of adjusting its frequency? I mean, most new gear uses 10MHz precisely.

[Conrad Hoffman]

Frequency counters and such tend to use 10 MHz, but it's very handy to have other frequencies available if you're aligning receivers or maybe even want to "fool" a 10 MHz device into reading with an offset.

[edpalmer42]

If you already have a good synthesized signal generator, you probably don't need Bodnar's variable-frequency GPSDO.  His units seem to be the best of the 'new breed' of GPSDOs that are aimed at hobbyists rather than professionals - although I suspect a lot of pros are using his devices instead of the more professional units with their much higher price tags!

The Z3805A and T'bolt units were top of the line in their day, but they're getting old.  The Tbolt has reached the end of its GPS rollover time so it now thinks the date is many years in the past.  I've heard of Z3801A units that have done the same, although mine is still hanging in.  I don't know how the Z3805A reacts.  Both Lady Heather and the NTPD package have been updated to compensate for the rollover in the Tbolt.  Of course, that isn't an issue if you're only interested in frequency.

Ed

[FriedMule]

What types is today more in the league with my units had at their time as new? Hmm maybe not the best formulation... what unit today is in same comparable quality to other new units as my units were to other time equal units?

[pdenisowski]

I'll take a look at the Leo Bodnar but what are the advantage of adjusting its frequency? I mean, most new gear uses 10MHz precisely.

I have the Leo Bodnar unit and would recommend it as well. I measured the phase noise performance using our flagship FSWP phase noise analyzer and the results were impressive for a GPSDO in that price range.

The frequency adjustment is useful if you're trying to replace a less-accurate, non-10 MHz clock.  For example, the ICOM 9700 amateur radio transceiver has an internal frequency reference of 49.152MHz, and by using an external GPSDO, you can significantly increase the frequency stability when transmitting / receiving on higher frequency bands (440 / 1200 MHz).  In fact, I would say most "weak signal" operators use the Bodnar box if operating on these bands.

I should also note that the Bodnar only goes down to 450 Hz - there's no PPS output (if that's important to you).

[bingo600]

For a COTS Unit.

My guess would be Ie. A Fury from Jackson Labs
https://www.jackson-labs.com/index.php/products/fury

For a " Semi DIY" prob. a Samsung (seems to use LEA-6T)
https://www.ebay.com/itm/284130731451
https://www.ebay.com/itm/284130731451

/Bingo

But the 10MHz from the TB or the HP - Are still "Some of the Best in class" (for used Telecom units) , even though the date is screwed up.
Just need a "Good antenna view" to the sky.
Ie. Due to the design - The TB has no "Hanging bridges"

[pdenisowski]

If you already have a good synthesized signal generator, you probably don't need Bodnar's variable-frequency GPSDO.  His units seem to be the best of the 'new breed' of GPSDOs that are aimed at hobbyists rather than professionals - although I suspect a lot of pros are using his devices instead of the more professional units with their much higher price tags!

I would agree with this statement - if you have a modern RF sig gen.  The advantages of the Bodnar GPSDO are small size, lower power consumption and much lower cost, so much more appropriate for the hobbyist market.

The reason I say modern is that I have some very old, non-R&S (I won't say whose) signal generators in my "lab."  When I was working on my phase noise videos, I wanted a signal source with somewhat poor phase noise performance to use in example measurement screenshots, and it turned out that in some cases, the Bodnar unit out-performed a few of the (professional) boat anchor sig gens I have. 

[FriedMule]

Money is of cause a problem or else could I just buy an atomic clock. :-)
My experience right now with my units are the accuracy fine and properly what I need,
but they do not agree 100% and that has made me to want a GPSDO that is of best for
a sane prize. The two old units, I have, were properly some of the best in their time.
Now would I like a comparable best unit of today.

All my gear uses 10MHz, frequency counter is a 12 digit type and I would love to be able to
know it is fairly accurate, so I am able to detect errors in my old GPSDO's. :-)

[807]

I'll take a look at the Leo Bodnar but what are the advantage of adjusting its frequency? I mean, most new gear uses 10MHz precisely.

I have the single output version of the Leo Bodnar unit. Thought I would use it to measure the frequency accuracy of the stations on the LW/MW bands. I could have set it to 10MHz to use as external reference for my Marconi 2022 sig gen, but thought I would play with the frequency settings of the unit directly. I screwed in an SMA to BNC socket adapter & poked a wire into it so my radio could pick it up.

I found out while doing this that not all frequencies can be generated with 1 Hz resolution. For example, Smooth Radio on 603kHz had a 1Hz beat. So I tried setting the unit to 602999Hz. It searched for 3 and a half minutes before declaring that it wasn't possible. BUT..it offered to generate a frequency of 602998.999814264Hz, less than 186 microherts difference. Good enough for me! 

btw...it normally finds a valid frequency in around 2 to 3 seconds. Then another 2 to 3 seconds to lock. That's using the antenna sitting on an inside windowsill.

[mendip_discovery]

Once you have one, better start dialling in the errors of your setup,
https://www.researchgate.net/publication/320798359_Evaluating_the_Frequency_and_Time_Uncertainty_of_GPS_Disciplined_Oscillators_and_Clocks

[bingo600]

Money is of cause a problem or else could I just buy an atomic clock. :-)
My experience right now with my units are the accuracy fine and properly what I need,
but they do not agree 100% and that has made me to want a GPSDO that is of best for
a sane prize. The two old units, I have, were properly some of the best in their time.
Now would I like a comparable best unit of today.

All my gear uses 10MHz, frequency counter is a 12 digit type and I would love to be able to
know it is fairly accurate, so I am able to detect errors in my old GPSDO's. :-)

 
A classic case of :
“A man with one clock knows what time it is, but a man with two is never sure.”

We all started that way .... 
Watch out - It's a slippery (&expensive) slope to walk.

Once you have your favorite GPSDO, you'll need to verify your Counter .... And so it begins.

/Bingo

[FriedMule]

To my usage would 186 micro z error be more than good enough, but for my personal fun and to make the foolish dive into the deep rabbit hole of accuracy and GPSDO would I love to get a unit that is considered great.

Maybe to explain it with another equipment I know a bit about:
You have the run-of-the-mill bench dmm's like the siglent 4.5 that is fine for most hobbyists, and you have the HP 8.5 that is in the top of accuracy and price!!!
Then do you have a unit like the Keithley DMM6500 6.5, it is fairly manageable, compared to the same as a 7.5. And still is the Keithley DMM6500, a dmm that outperforms most other dmm's in its class.

I would like to get a Keithley DMM6500 quality GPSDO:-)

[edpalmer42]

I don't know anything about today's crop of professional level GPSDOs so let's ask a few questions and see what happens.  All of these questions should be answered in terms of value to hobbyists rather than professional users.  It seems reasonable to me that current model units will almost always be more desirable to professional users.

First, let's list some obvious advantages to today's units.
- More sensitive receivers.
- More simultaneous channels.
- More GNSS systems.  i.e. GPS, Glonass, Beidou, Galileo, etc.
- Faster cold or warm start from power-on to GNSS lock

1.  Do today's professional-level GPSDOs provide a more accurate 10 MHz frequency output than older professional level GPSDOs like the Z3801A and Tbolt?

2.  Do today's units provide a 1 PPS output with less jitter than older units?

3.  Do today's units produce an output with better Allan Deviation than older units?

4.  Do today's units provide other features that make them clearly superior to older units?

-----------  end of questions ---------
----------- beginning of random rambling -----------------

A single-frequency, i.e. L1 at 1575.42 MHz, GPS can resolve a location to an accuracy of a meter or two.  Most of the error is caused by distortion to the signal caused by the ionosphere. Some of the error is due to poor geometry related to the orbit of the satellites and the location of the antenna on the ground.  A dual-frequency, i.e. L1 and L2 at 1227.60 Mhz, unit can resolve to a few centimeters or better and is typically used for land surveying.  Does anyone make a dual-frequency GPSDO?  Would the navigation improvements translate to timing and/or jitter improvements?

Ed

[pdenisowski]

A dual-frequency, i.e. L1 and L2 at 1227.60 Mhz, unit can resolve to a few centimeters or better and is typically used for land surveying.  Does anyone make a dual-frequency GPSDO?  Would the navigation improvements translate to timing and/or jitter improvements?

For the time being, most dual frequency civilian receivers use L1 and L5 (not L2).  Some of the newer u-blox and Trimble chips do provide higher timing accuracy via dual band operation.

https://content.u-blox.com/sites/default/files/documents/NEO-F10T_ProductSummary_UBX-22025534_0.pdf
https://timing.trimble.com/wp-content/uploads/Datasheet_RES720-v1.pdf

Not sure if these fall into the "hobbyist" category in terms of price, but they do indicate that dual-band could be used to increase timing accuracy.

[edpalmer42]

A dual-frequency, i.e. L1 and L2 at 1227.60 Mhz, unit can resolve to a few centimeters or better and is typically used for land surveying.  Does anyone make a dual-frequency GPSDO?  Would the navigation improvements translate to timing and/or jitter improvements?

For the time being, most dual frequency civilian receivers use L1 and L5 (not L2).  Some of the newer u-blox and Trimble chips do provide higher timing accuracy via dual band operation.

I didn't realize that the pros had moved to L5.  Older survey receivers were all L1/L2.

https://content.u-blox.com/sites/default/files/documents/NEO-F10T_ProductSummary_UBX-22025534_0.pdf
https://timing.trimble.com/wp-content/uploads/Datasheet_RES720-v1.pdf

Not sure if these fall into the "hobbyist" category in terms of price, but they do indicate that dual-band could be used to increase timing accuracy.

I thought that if dual band would help, somebody would be doing it.  But since the NEO-F10T was announced less than 2 weeks ago I don't think they'll be available to mere mortals for some time.  The RES720 is available, but it doesn't have a 10 MHz output which makes it more annoying to use in a GPSDO.  However, the ICM720 is another version of the RES720 and does have a 10 MHz output, but I can't find a price for it.     The RES720 is available for a somewhat reasonable 192 euros so hopefully the ICM720 is similar.  Note that all of these units are bare GPS receivers, not GPSDOs.  It's interesting that both the Trimble and u-Blox units spec jitter as 5 ns (1 sigma).  Back in the late 2000s, I picked up a Navsync (now Connor-Winfield) CW-12 GPS Receiver (not a GPSDO) that spec'ed 30 ns with 5 ns resolution.  My measurements showed 4 - 5 ns (1 sigma) with a range of 20 - 25 ns.

FYI, here's some old measurements I took of some bare GPS receivers and GPSDOs.  Note that some people who know far more than I do about this stuff have told me that these measurements don't come close to adequately measuring the quality of the units.   

Code: [Select]

GPS Devices -- Measure & analyze the period of the 1 PPS Output

Device ............... Std Dev (ns).... Range (max-min)(ns) ... Device ...... Notes

Navsync CW12 ......... 4 - 5 .......... 20 - 25 ............... GPS Rcvr .... 1,7
Motorola UT+ ......... 40 - 55 ........ 95 - 110 .............. GPS Rcvr .... 2,7
Rockwell Jupiter ..... 10 ............  50 ...................  GPS Rcvr .... 3,7
Motorola M12M ........ 10 - 15 ........ 40 - 60 ............... GPS Rcvr .... 7

Trimble Thunderbolt .. 0.4 - 0.5 ...... 2 - 4 ................. GPSDO ....... 6,8
HP Z3801A ............ 0.1 - 0.2 ...... < 1 ................... GPSDO ....... 6
HP Z3817A / CW12 ..... < 0.1 .......... < 1 ................... GPSDO ....... 4,6
Jackson Labs GPSTCXO . 0.3 - 0.4 ...... 2 - 3 ................. GPSDO ....... 6
NEC NWM-034241-201 ... 0.1 - 0.2 ...... < 2 ................... GPSDO ....... 5
Trimble UCCM ......... 0.09 - 0.11 .... < 1 ................... GPSDO ....... 5

Results are based on multiple runs of ~ 1000 measurements each.
Sawtooth correction has not been used for any of the GPS receivers.  Where supported, it would reduce the numbers substantially.
All units were connected to the same antenna system.

Notes

1.  Sawtooth correction not supported.
2.  Most 'range' results were in this group, but there were a few at 20 - 30.
3.  Only one test.
4.  Requires external 1 PPS input.  Equipped with E1938 oscillator.
5.  Measurement made with Fluke PM6681.
6.  Measurement made with HP 5370B.
7.  Measurement made with HP 5372A.
8.  Standard parameters.

Ed

[pdenisowski]

I didn't realize that the pros had moved to L5.  Older survey receivers were all L1/L2.

Actually, the "pros" (like surveyors) still do use L2.  The issue with using the civilian codes on L2 is that L2 is not a "protected band" so it can't be used for things like civilian aviation, safety of life, etc.  L5 is in a protected band, and that's one of the main reasons why a lot of dual band receivers now use L5.  I mention this in my video on GPS links and codes

https://youtu.be/_VJZQMGEG1k?t=572

Using L5 also helps with ionospheric corrections because it's farther away in frequency from L1 than L2, and greater frequency separation yields better ionospheric corrections for dual link receivers.

[edpalmer42]

Thanks for the clarification and the video link.  That's much more info than most time-nuts need, but it was very interesting.

I've kind of lost interest in GPSDOs over the last few years because the performance of all the units is basically defined by GPS and, as a result, is both similar to each other and 'good enough'.  Better oscillators help with short term results, but you quickly run into diminishing returns there.  Other features are, for me, mostly nice to have, but not significant.  But these dual band units could open up new possibilities.

Ed

[mendip_discovery]

I've kind of lost interest in GPSDOs over the last few years because the performance of all the units is basically defined by GPS and, as a result, is both similar to each other and 'good enough'.  Better oscillators help with short term results, but you quickly run into diminishing returns there.  Other features are, for me, mostly nice to have, but not significant.  But these dual band units could open up new possibilities.

Ed

Don't let the time loons hear you talk like that. Otherwise they will make you do the Time Warp. It just a jump to the left...

[MIS42N]

I am interested to read about GPSDOs. Rarely do people state what they want to use it for, how accurate it needs to be, will it be on continuously or turned on when needed, where is the antenna likely to be located. I have been experimenting with various very cheap configurations and the best figures are like - worst deviation of OCXO 10MHz to GPS 1pps is ±50ns or half a cycle, worst deviation from 10MHz is ±1mHz. And because it is locked to the GPS signal, will long term be accurate. Just changing the OCXO these figures can be improved. Using a dual frequency GPS will also improve it. More attention to circuitry will improve it. I have been reading about Johnny B Good and his attempts to eliminate diurnal fluctuations from his Rubidium standard, with the conclusion the deviations are most likely due to the GPS signal, not the Rb. I don't believe JBG has a reason to do this other than pursuing perfection. He may have to buy a dual frequency GPS, the only justification being he can.

Some people do have a genuine need. Long baseline interferometry needs accurate time. Radio transmissions in the 10s of GHz need accurate frequency. The figures I gave above would probably satisfy someone broadcasting at 10GHz, with an error of 1Hz. But the LBI requirement could be less than 5ns, so 50ns won't be adequate.

So without a requirement or a budget, answering 'What GPSDO do you recommend' is nigh impossible.

[max-bit]

Trimble THUNDERBOLT GPS

Professional product...
I've been using it for years