Review Jcat Master OCXO Clock – Don’t Be Square!

Measurements and conclusion

We did some measurements of the Master Clock as well as the USB XE. We did these on the Wavecrest SIA 3000. The Master Clock was connected directly with a coaxial cable. The USB XE was measured with an active probe, as it does not have an SMA connection.

Most of the measurements on the Master Clock are in the noise floor of the Wavecrest. (Which is around 2.4ps jitter). However, it is still interesting to see what the noise does when we switch power supplies. So keep in mind that with the Master Clock we are at the limits of the Wavecrest in terms of jitter.

Phase Noise on the new Aeroflex we have not been able to measure because we do not have a 20 MHz REF clock. We are saving up for a signal generator with low phase noise. However, those are very pricey. So that may take some time.

Measurements Master Clock

We powered the Master Clock with three power supplies: a switchmode power supply, a Pura Dodo and a Sbooster BOTW 5V.

The measurements show that Jcat stabilizes the power supply well before it goes to the clock. There are differences, but they are minimal. Even with the PC’s power supply (we’ll see later), the difference is very small. This is positive, of course. We see no problem in powering the module with a properly filtered, internal solution. You don’t really need a very expensive external DC power supply.

Measurements USB XE clock

The USB XE certainly has a decent crystal, but it is nowhere near the Master Clock. Where we are at the limit of the Wavecrest SIA with the Master Clock, here we see values around 12ps. Certainly not bad for a clock on a USB card of course, but it is not the level of the Master Clock.

We also see that the power supply has more impact. The Dodo is the cleanest, the Sbooster is not a good match, as we see more jitter compared to the switchmode power supply as well as the Dodo. Strange though; we didn’t expect that. By the way, we do see the noise increase with the switchmode compared to the Sbooster. At the bottom line, however, we see that the Dodo is the quietest solution in all areas.

Measurements Master Clock on USB XE

What we see as soon as we connect the USB XE (with the external clock input) to the clock is that the noise as well as jitter increases slightly. It seems almost impossible for us to avoid this. We made a combination of both the Master Clock and the USB card to the internal computer power supply, both to an Sbooster power supply as well as the USB XE to an Sbooster. We did not have the Dodo available for a while during this measurement.

Bizarrely, it works better with both devices on the internal power supply. Now we have already seen that the Sbooster does not work very well in this combination. We estimate – due to the previous measurements – that it drops (slightly) with the Dodo.

Conclusion

Technically, the Jcat Master Clock and USB XE is a very solid product. Influence of power supply is small if we look purely at jitter and noise. Sound-wise, it can be bigger, because noise from the USB port to the DAC can also flow. We definitely hear differences when we change power supplies, although of course it is and remains within margin and the influence depends on the integration.