Compare GPSDO LBE-1420 vs Mini vs OCXO 10MHz

[KedasProbe]

Hi,

I have the Leo Bodnar LBE-1420 GPSDO and The older Mini version.
So I though I should compare then.
Since I have only two references, not knowing which one is changing, I added an extra OCXO reference.
(No, I don't have a Rubidium atomic clock)

I use the TinyPFA to compare the two GPSDOs
Since I compare to a similar product and not a better reference I assume the measured error is about √2 too high. (not sure don't know the appropriate math, assuming both devices are very similar)

For the OCXO I used a cheap one from Amazon 13 Euro and fine tuned it to the GPSDO. (OSC5A2B02)
I used the sine wave output, all 3 devices are loaded with 50 Ohm.
I did notice a problem with putting a load on the TTL output, the voltage on R1 increases 2mV, the freq changed about 62mHz and the amplitude of the since wave dropped almost in half.
So you may want to check that before using it. (I don't care about that second output)
Attached is an DMM6500 image of the R1 voltage when I attach a 50ohm termination to the TTL output.

Power Usage:
Mini: 0.7W (set to 8mA)
LBE-1420: 1.2W  (a bit high compared to a OCXO, the housing does get warm after a while, no output power setting like the Mini)
OCXO: 1.4W but will increase if you increase the power supply voltage 7-13V. (+extra power adapter)

I keep the Mini powered and connected to my function generator, I like the lower power of that one.
The LBE-1420 does have a more accurate (stable) time representation in its UI than the Mini.
Although the LBE-1420 can go as low as 1Hz it's not an 1PPS (random phase), not sure if that is a hardware or software limitation. 1PPS should be possible with the to be released newer dual channel version.

TinyPFA:
To connect the signals to the TinyPFA I added -20dB on the LBE-1420 output and -15dB on the Mini.
Attached you find some graphs and the .tim data files of Timelab. https://www.miles.io/TimeLab/beta.htm You can open the file in that program. (one is the measured noise floor to verify its measure limits.)

I also made a scope recording since that can give a easier to understand impression of the stability.
https://youtu.be/Iy8IpG-3oGU
Check the average phase difference on the right, an 'Dev' that goes up means it changed 'fast'

Note 1: Showing 3 channels at the same time wasn't possible due to the sample rate limitations of this scope. (results in unstable phase measurement)
Note 2: I wanted to make a longer recording and speed it up but the Scope app keeps crashing when I try downloading long recordings.

[EvgenyG]

The LBE-1420 may be good, but mine stopped receiving GPS signals some time ago. It was just sitting on my desk not even connected to anything. Raised this with support, was advised to check the antenna. I've ordered 2x different ones from Ebay, one of them arrived and made no difference. I will get another one and will try again. This started happening after firmware upgrade to 1.05.

[Leo Bodnar]

Hi Evgeny,
We are sending you a new one and it would be great to get the faulty one back to see what went wrong there.
Sorry for the trouble and thank you for your patience.
Leo

The LBE-1420 may be good, but mine stopped receiving GPS signals some time ago. It was just sitting on my desk not even connected to anything. Raised this with support, was advised to check the antenna. I've ordered 2x different ones from Ebay, one of them arrived and made no difference. I will get another one and will try again. This started happening after firmware upgrade to 1.05.

[EvgenyG]

Thanks Leo, new unit received and works great. Faulty unit posted back, your support team has tracking details. I'd be curious to find out what's wrong with it, too. The new unit seems to be picking up more satellites/getting fix quicker than the previous unit did. I am seeing about 50-70% of the bars showing reception, but the other unit showed only 6-7 satellites (~30% of bars) before it stopped receiving at all.

[GigaJoe]

OSC5A2B02  -- 5V nominal ,  not sure how long it will last under 12 V
usually   oscillator takes around a week up to month , to settle

to minimize jitter need additionally isolate metal case from the air ,  to minimize temp fluctuation, Styrofoam box , enclosure ... etc
Adjustment voltage on pin need very stable, clean and filtered, this particular oscillator,   sensing around 0.2-0.5 mV, not the best for adjustment.   other usually 0.1mV ,  so your reference voltage as source of V adj. ; should have stability better then 0.1mV. (just a note Vref 431 kind , may drift up to 10mV ).

[KedasProbe]

Some Timelab data (zip file) from the LBE-1420 GPSDO on the GB7TBL FA-5-6GP, I added -3dB.
(using internal ref after call on 10Mhz CH1)

How much time does the LBE-1420 need to change to a stable 1GHz from 10MHz?
Something changed slow (5min.) but not sure what, could be the FA-5. (see pic)

Note: haven't found a way yet to change the FA-5 gate time, always 0.1sec.