Discrete Low Jitter Clock GB

[amc184]

No, you can't skip the comparator if you want a clock output that is suitable as a master clock for a CD player (the purpose of the project). You could replace its functionality with discrete components, but it's not something I'd recommend.

[wendell c]

any more pcb's available?

[amc184]

Yes, I've still got a couple of PCBs spare, PM me your details.

[andrea_mori]

Any pcb's available?

[korben69]

Hi Anton
I've finished one LJC module today.
Measuring 4.93V on pin1 of AD8561 vs 5V.
Is it Ok ?
What kind of wire do you use ?
Thanks

[roender]

Did someone measured the phase noise in frequency domain for this clock?

[studiostevus]

Originally Posted by amc184

No, you can't skip the comparator if you want a clock output that is suitable as a master clock for a CD player (the purpose of the project). You could replace its functionality with discrete components, but it's not something I'd recommend.

Can you please elaborate?
What's the function of the comparator? Purely buffering?

thanks

[amc184]

@korben69: That's fine, well within margin. The output connector pattern suits SMA / SMB / SMC connectors, so usually you'd use something like RG316. The cabling isn't critical over very short distances, and the clock should be positioned very close to the point where it is to be used so that the transmission distance is kept very short. The clock isn't designed to drive a transmission line. Sorry about the late reply, I hope the answer is still useful.

@roender: No, I've already answered this question.

@studiostevus: The comparator is absolutely essential, it can't be omitted. Its function is to transform the two out of phase sinusoidal signals from the oscillator into a single squarewave. It's not merely a buffer.

[studiostevus]

Thanks for explaining.

Any idea about the effect of the comparator on jitter/phase noise...? The data sheet mentions a 7ns propagation delay, no mention of jitter (15ps typical i read somewhere). In any case, i believe that any psu noise to the comparator will result in phase noise...

Worth mentioning is also that not all devices require a square waveform to trigger. Have you tried to use it without?

[amc184]

The comparator will add some phase noise, but it's not separable from the rest of the design. However, what it adds is by far outweighed by what it does by providing a squared output and the cancellation of common mode noise from the oscillator.

In any case, i believe that any psu noise to the comparator will result in phase noise...

Did you note that the comparator runs off a separate regulated supply to the oscillator?

Worth mentioning is also that not all devices require a square waveform to trigger. Have you tried to use it without?

As long as the waveform is of adequate voltage, almost any wave shape (triangle, sine etc) will work with almost any device. However, one of the significant causes of jitter is noise in the transition voltage of the receiving device. The jitter caused by this is proportional to the slope of the waveform about the transition, the steeper the slope the less jitter. Why would you want to use anything but a square wave?