[bodiebill2]

I am sure this has been asked and answered prior but I am looking for opinions to why. I am using the LHY OCK-2 clock with my RS130 (transport streamer) and Audio-GD R7HEMk3 (DAC) both connected via 50ohm BNC to the LHY OCK-2 on the 50 ohms sine ports. I started off with Square wave and have not tried Sine Wave until today. To my surprise Sine sounds better and my question is why. I thought Square wave was superior to Sine. I am no engineer so my assumptions could be wildly off. Any insight would be appreciated.

Square theoretically is better, but it depends on the implementation/quality. Also, square requires top notch cables with the right impedance. So with lower quality or mismatched cables sine can sound better.

By the way I am using a Mutec Ref10 SE120 (square only) to feed a Rose RS130 and DI20HE (both 75 Ohm).

 

[sajunky]

I am sure this has been asked and answered prior but I am looking for opinions to why. I am using the LHY OCK-2 clock with my RS130 (transport streamer) and Audio-GD R7HEMk3 (DAC) both connected via 50ohm BNC to the LHY OCK-2 on the 50 ohms sine ports. I started off with Square wave and have not tried Sine Wave until today. To my surprise Sine sounds better and my question is why. I thought Square wave was superior to Sine. I am no engineer so my assumptions could be wildly off. Any insight would be appreciated.

Do you use 'isolated' option with a 'square' output?

I see there are different opinions which is better with the OCK-2. I have a comment for this particular "OCK-2" case.

A non-uniform shape of a squarewave is not a problem with a direct connection. It can still deliver accurate timing. However problem arise when using isolated option. In such case unregular shape can cause false triggers, as the output is not exact DC voltage shape, but a derivative (du/dt).

Not having "OCK-2" to test, but based on the above, the 'isolated' position of the switch is not an option with a square port. I can be wrong, but keep it in mind.

 

[duffer5]

Do you use 'isolated' option with a 'square' output?

I see there are different opinions which is better with the OCK-2. I have a comment for this particular "OCK-2" case.

A non-uniform shape of a squarewave is not a problem with a direct connection. It can still deliver accurate timing. However problem arise when using isolated option. In such case unregular shape can cause false triggers, as the output is not exact DC voltage shape, but a derivative (du/dt).

Not having "OCK-2" to test, but based on the above, the 'isolated' position of the switch is not an option with a square port. I can be wrong, but keep it in mind.

Pardon my ignorance but what do you mean by isolation option.

 

[sajunky]

Pardon my ignorance but what do you mean by isolation option.

There are two dip switches for each port.

- 50/75 and
- ISO (means isolated)

When ISO switch is set to 'on' position, port is isolated.

 

[duffer5]

There are two dip switches for each port.

- 50/75 and
- ISO (means isolated)

When ISO switch is set to 'on' position, port is isolated.

Actually I have yet to open up the case. What is the benefit of the ISO setting. Thank you for educating me. I guess I should know these things.

 

[Superdad]

There are two dip switches for each port.
- 50/75 and
- ISO (means isolated)
When ISO switch is set to 'on' position, port is isolated.

Well not really isolated. Just a capacitor across the separated ground domains. It does block DC and filter some AC. It won't block all leakage current, and of course it has to pass the 10MHz clock signal.

Compare to full, active driver/isolation circuit and transformer on each output of Mutec:

 

[duffer5]

Independent to the impedance setting. ISO means "isolated", it gives isolation for DC (including mains 50/60Hz and low frequency harmonics). This is setting for protecting from ground loops. GND position gives a direct ground connection with the receiver.

Use settings which works the best for you, it is system dependent. If you have two or more receiving devices (like DDC and a DAC), I would suggest using GND option only on the one connection (for DAC in my preference). It may conflict with preferences for square and sine outputs, I have no experience in this area, as my RF experience is limited to the laboratory equipment and HAM radio.

@sajunky I found your prior answer as I am combing through this mega thread. Thanks for the ongoing support.

 

[sajunky]

Got it and thank you for the explanation. I do not have the other ports isolated. I assume if I had the 4 non-used ports in ISO mode this would be an audible benefit? I am only using two ports at the moment. One for my streamer and one for the DAC.

Keep all unused ports non isolated. For the sine port you can use both. For the square port isolated option is problably not a good idea. It is based on these published scope screenshots. Not sure this source can be trusted, just to be safe. If slope transition were smooth, I would have no reservation.

How to chose which port should be isolated? When clock serves more than one device, I would chose non-isolated port for a DDC (or a DAC if there is no DDC) and all other connections isolated. This is from engineering point of view, as a general guideline.

 

[sajunky]

Well not really isolated. Just a capacitor across the separated ground domains. It does block DC and filter some AC. It won't block all leakage current, and of course it has to pass the 10MHz clock signal.




Compare to full, active driver/isolation circuit and transformer on each output of Mutec:

It looks the same (circled area) if you look close. Follow the arrows:

 

[duffer5]

Thank you, @Superdad and @sajunky for the replies, super helpful.

 

[Superdad]

It looks the same (circled area) if you look close. Follow the arrows:

Not the same at all. Your arrows are pointing to tiny custom transformers. The photo is just not high enough resolution to see that.
And there is the whole multi-stage active buffer circuit before each one.

 

[sajunky]

Not the same at all. Your arrows are pointing to tiny custom transformers.

Clearly shown on this photo:

 

[Superdad]

Clearly shown on this photo:

You are pointing to common capacitors. The custom transformers are just below that.
Since you reported me and had my post of facts deleted, I am done here.

 

[sajunky]

You are pointing to common capacitors

Of course a capacitor. The same critical component which you put in a circle on the OCK-2 suggesting there was something wrong.

 

[Superdad]

Of course a capacitor. The same critical component which you put in a circle on the OCK-2 suggesting there was something wrong.

I did not. Just pointed out that switching in/out a capacitor across the separated ground domains does not constitute actual isolation. As I said, it does block DC and filter some AC. But it won't block all leakage current, and of course it has to pass the 10MHz clock signal.