What GPSDO do you recommend?

[MIS42N]

Trimble THUNDERBOLT GPS

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

I thank the author for posting this, but it illustrates my post. What does the author do with their GPSDO? What accuracy is required? mobile or fixed? type of antenna? new or second hand? use the 1pps? interested in the NMEA data? was there a cost/benefit analysis that picked this out from other contenders and what criteria were used if there was?

So many questions, so few answers.

[J-R]

My 2 cents: I picked up the Leo Bodnar unit to have something to check/calibrate my function generators and frequency counters and also as a possible 10MHz standard for devices with an external clock input.  None of them go beyond about 10 digits, so I don't think I would classify any of it as metrology-grade.  I didn't want the ebay specials, and I didn't want to spend $1k, so the dual-output Leo Bodnar unit hit the sweet spot for me personally.

It is handy to be able to generate other frequencies with high precision, even if nothing else to sanity check devices at the full range of supported frequencies.

However, due to the slightly clunky software and minor limitations of the dual-frequency model, I would not suggest using it on a daily basis in your workflow to generate a lot of different frequencies quickly.

Ultimately I ended up not using it permanently as a distributed 10MHz standard, as all of the equipment I have is good enough without it, and furthermore I didn't like the idea of trusting a single source.  So if things disagree later, I check check into it at that time.

[807]

Something I found a bit odd about the Leo Bodnar, was that in my earlier post, I wanted a frequency of 602999Hz. It couldn't produce it, but came up with something very close (602998.999814264Hz). Knowing that it couldn't output 602999, when I wanted to generate that frequency again, I put in 602998.999814264Hz directly...but after waiting for the unit to think about it, it said that frequency wasn't available! Instead it suggested the exact same frequency. 

[mendip_discovery]

Something I found a bit odd about the Leo Bodnar, was that in my earlier post, I wanted a frequency of 602999Hz. It couldn't produce it, but came up with something very close (602998.999814264Hz). Knowing that it couldn't output 602999, when I wanted to generate that frequency again, I put in 602998.999814264Hz directly...but after waiting for the unit to think about it, it said that frequency wasn't available! Instead it suggested the exact same frequency. 

I suspect a floating point/rounding issue. I am surprised it decides to go for so many digits as that looks to be in the jitter zone. Leo might be able to correct it in the software if you message him about the issue.

[J-R]

I had no trouble generating 602999Hz on my dual frequency unit.

[MIS42N]

I had no trouble generating 602999Hz on my dual frequency unit.

It could have been 602998.9998Hz. You are unlikely to know unless you have a real 602999Hz to compare, and can wait a minute or two to note the slight divergence. Most hobbyists would not worry about a difference of less than 1 part in a billion. And, over a minute, is the Leo Bodnar accurate to 1 part in a billion or off chasing some ionospheric disturbance that disturbs the GPS 1pps for a short period?

I note the byline " Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification" - What is the Leo Bodnar specification and has it been verified?

My design differs from the standard PLL by making periodic (usually about every 7 minutes) rather than continuous adjustment to the control voltage of the local oscillator. That allows the unit to be self verifying. I haven't seen that in any other DIY unit.

[J-R]

Not sure what your point is, it was mentioned that the single output unit was not able to accept 602999Hz as an output setting.  I'm simply stating I did not have the same issue.

[MIS42N]

Not sure what your point is, it was mentioned that the single output unit was not able to accept 602999Hz as an output setting.  I'm simply stating I did not have the same issue.

My maybe not clearly stated point is - maybe the hardware is the same as the person who couldn't get an exact 602999Hz, but something that was very close. Update the software so if it is asked for a particular frequency, it just dishes up the closest it can without alerting you to the fact that it isn't exactly what you asked for [within reason - if it couldn't get within a few ppb then alert]. Most people would not notice or care about the difference.

[J-R]

The issue being discussed is that when you request a specific frequency it has to do some gymnastics before accepting it.  If it can't provide that frequency, you will get an error or it may just sit there searching endlessly without a lock.  I've only ever seen it when requesting very odd frequencies well past the decimal point, say 152,384,577.4234195Hz or etc.

[4thDoctorWhoFan]

I now have two different GPSDO units and they agree with each other down to 1mhz.  This is good enough for me.

One is the BG7TBL type Chinese made GPSDO with the 2 line blue LCD screen which I bought about 3 years ago and the other is the Leo Bodnar made unit with the dual output which I have had for about a year.  I leave them on 24/7 and they agree with each other down to 1 milli hertz.

I only purchased the Leo Bodnar model for a sanity check and I'm glad I did. 

[Zenwizard]

For me long term stability is a Samsung GPSDO though I also have the Leo. I use the Leo for other things in the lab. For short term I have a Stanford research 725 fed by the 1pps from the GPSDO. But I have also been asked to help characterize GPSDO's and smash them in to the Rb to get some drift measurements from the limmited testing the leo and the samsng both barely break 100uHz of sdev over the course of a few hours. So it is REALLY getting in to the weeds.