LBE-1420 "infinite resolution" GPS-locked any frequency between 1 Hz and 1.6 GHz

[bb1]

Leo Bodnar GPS disciplined clock source LBE-1420 description at www.leobodnar.com claims
GPS guaranteed relative output frequency stability of at least 10^(-11) (0.01 part per billion (ppb)).
Very low phase noise.
Any output frequency between 1 Hz and 1.1 GHz can be set with "infinite resolution".

It is difficult to verify "infinite resolution" claim, but I can confirm at least 10 ppb (10^(-8)) resolution.

I calibrated Tek 2712 spectrum analyzer counter at 1.000 000 000 GHz, and then used Leo Bodnar software to set LBE-1420 output frequency to
1.234567890 GHz.
To protect 2712 I used 20db attenuator, so the signal strength on the 2712 screenshots is 20 db below the actual LBE-1420 output level,
which, according to Leo Bodnar, is 11db at 10 MHz.

As can be seen on the attached 1.234567890GHz_rbw1kHZ_scr.jpg, the Tek 2712 counter measured 1.234567897 GHz,
with obvious error of 7 Hz (7 ppb) at 1.2 GHz.

On this figure one also can see the very clean spectrum peak, no any kind of spurs around.

While officially the top frequency of LBE-1420 is claimed to be 1.1 GHz, the software limits actual output frequency to 1.6 GHz.
Spectrum at this frequency is shown on 1.6GHz_rbw9kHz_scr.jpg.

While low limit of the output frequency is 1 Hz, the frequency resolution near this limit is also good. Keysight scope screenshot
ks_1.2345Hz_scr.png shows that the frequency can be set to 1.2345Hz. Here the input is not 50 ohm terminated.
More accurate measurement of the frequency 1.23456789 Hz can be done by frame time differences in sequence mode, which have about 1 ns uncertainty.
It also confirms at least 10 ppb LBE-1420 frequency resolution.

Many instruments have 10 MHz external frequency input. The 2712 screenshot at this frequency is shown on
10MHz_span20MHz_scr.jpg.
2712 here has 1 Hz error due to 1 Hz 2712 frequency counter resolution.
The spectrum 10 MHz below and above central frequency is very clean, no spurs.

Finally, it is interesting to see the shape of the LBE-1420 output signal at higher frequency of 99 MHz with 50 ohm termination.
1GHz Tek scope screen capture is at ts_99.99MHz_rise_fall_av100_scr.png.
Tek clock is not very accurate.
One can see the rise and fall times of about 400 ps.
As the scope official rise time is 450 ps, the LBE-1420 rise/fall times must be much lower.
Leo Bodnar claims rise/fall times 300/260 ps at 10 MHz.

[drew23]

Hi.

I have tested the phase noise of the LBE-1420 and got much the same result as the specifications. Impressive.

It is a good price for an accurate and low-phase noise GPSDO signal generator.

Has anyone done a teardown or have an idea of the design?

Regards Drew VK4ZXI

[EvgenyG]

They've done a great job, I already used it to align/check some of my radio equipment and it worked great. I don't have access to expensive RF test equipment and now I don't need, because this little box does it all. As pointed out by bb1, square waves have pretty decent rise/fall times, can be used to check your scope bandwidth, too. Highly stable 100Mhz square wave? Easy. It can go even higher. For the price/performance you won't find anything better.