I used a cheap cheap clock

[Sunsun22]

I purchased a clock kit from China at around US$15 + US$1.5 courier charge to Hong Kong that includes PCB, all components, a "so to say" 10 ppm oscillator and a 2VA transformer. After assembling, I measured the output voltage and it reads 2.5V but my oscilloscope reads overshooting. I am fighting if I should put it in my antique - Pioneer T-07A which has been in my service since 1990. Eventually, I used 3 hours to install it.

When started, it already sounds excellent - the low is having quicker response/punch, the mid is getting sweet and there are many more details that I never heard before. After running in for about 5 days, it even turns better.

My questions: -

1. What causes such overshoot?
2. Does the overshoot hurts?
3. How to kill the overshoot?
4. They have a so call 3 ppm oscillator that will cost me US$19 extra. Does it affect the sound quality in great extend by changing the 10 ppm oscillator to 3 ppm?
5. The circuit is very simple. Why there are so many clocks in the market that has a lot of components? How good are they?

Sunny

[Sunsun22]

This is the cheap cheap clock. I have erased the brand name.

[aaronboumans]

nice! can you give me the name? or website?

Is your probe calibrated correctly?

[fmak]

My questions: -

1. What causes such overshoot?
2. Does the overshoot hurts?
3. How to kill the overshoot?
4. They have a so call 3 ppm oscillator that will cost me US$19 extra. Does it affect the sound quality in great extend by changing the 10 ppm oscillator to 3 ppm?
5. The circuit is very simple. Why there are so many clocks in the market that has a lot of components? How good are they?

-----------------------------------------------------------------------------------
You need to ensure that the clock output impedance is matched to the input impedance; try terminating into a 50R or 75R load.
You can kill some of the overshoot by mopunting a small capacitance across the output. Apparently some XOs like a bit of capacitance say 10 to 47 pf.

The clock looks quite well made. Can you give me details and I may order one myself.

Clock accurracy is not important. Low phase noise (jitter) is.

[davidsrsb]

The suspension made me chuckle, so little mass so the resonant frequency is going to be quite high.

[ash_dac]

Originally posted by Sunsun22

5. The circuit is very simple. Why there are so many clocks in the market that has a lot of components? How good are they?

Sunny

low noise psu >>> low phase noise sine wave oscillator >>> sine to square conversion

The extra parts go into the 1st and last bits.

Biggest problem will be layout so need a pcb!

I have yet to try feeding the sine wave directly into a decoder chip.

[poynton]

Originally posted by davidsrsb
The suspension made me chuckle, so little mass so the resonant frequency is going to be quite high.

Hi.

Agreed but it shows that some thought has gone into the design.
It does look well made. If we had some conponent values and a shot of the underside, we could reverse engineer the circuit.

Andy

[martin clark]

I have yet to try feeding the sine wave directly into a decoder chip.

Works fine so long as it still meets the voltage swing expected by the input logic. Don't worry about the apparently 'slow' rise time, that matters less than the cycle-to cyclce variance a poorly-considered inverter will add(!).

Subjectively it worked quite well IME too - i.e. building a clock and coupling the sine output into a couple of cd players sounded better than using the original, and no discernable improvement came from then adding-in an inverter to square-up the signal. But then - that's also adding the inverter's sensitivity to PSU noise into the equation too...

[georgehifi]

I've seen this before and had saved the circuit diagram for it .
enjoy

Cheers George

[Sunsun22]

Thank you for all the input. The measurement was made without loading. I will measure it when it is in operations. Yet, I still wonder should I install a more expensive clock in an attempt to even improve better?