Help! Kwak Clock - Partial adoption

[overmind]

I have constructed this clock, as attached, for my CD player. It is based on Kwak Clock, without the power supply as I only have a single ended power supply in my CDP.

Initailly, the circuit is based on KC6. Both the Jfet and AD8561 shared the same power supply of +5V (+V1). I have noticed that with +5V supply the output of AD8561 (clock signal) peaks at about 3.8V. I decided to try out if a clock signal of higher amplitude would make a difference. So I added in the additional circuit to give +V2 of 5.9V specially for the chip AD8561.

In my orginal construction, R2 is a metal film resistor. I decided to changed it carbon film. The rest of the resistors are all carbon film. R1 remains at 270ohm.

After done all these, I cannot detect a oscillation from the circuit, using a osilloscope. The voltage at Drain of the Jfet is 4.9V. Voltage at pin 1 of AD8561 measured to be 5.9V as well. I have tried to isolate the Jfet oscillator circuit from the AD8561 circuit + powersupply. Resistor R1 was also changed to 1kohm.

I have changed 3 pieces of J309, and still have not idea what went wrong. The circuit was orginally working.

I wonder would there be a possibility that the Crystal oscillator actually failed.

Please advice.
Thanks
SW

[alfsch]

Hi,
test: voltage on source/r1 ?
+ tested another crystal?
alf

[overmind]

Originally posted by alfsch
Hi,
test: voltage on source/r1 ?
+ tested another crystal?
alf

Tested around the parts of the oscillator circuit, could not pick up any signal, only +5V or 0V.

I only have one piece of 5ppm crystal, will try the one pulled out from my CDP.

Regards,
SW

[Lars Clausen]

I would exchange the 39p and 68p in the oscillator to some additional gain.

[Bricolo]

Originally posted by Lars Clausen
I would exchange the 39p and 68p in the oscillator to some additional gain.

are those values critical?
I can't find 39p and 68p here Also no NPO

[alfsch]

Hi,
at source/r1 you must get some voltage, 2V or so, otherwise this shows, the fet draws no current --> no osz... check fet connections, or try a new one, prob you killed it.
alf

[Lars Clausen]

Yes in the oscillator they actually form a sort of transformer, with the lower capacitor's charge as 'primary' and the higher (geographically on the sch) capacitor's charge as 'secondary'.
Please note that since the input charge equals the output charge, the transformed voltage is higher with lower capacitance.
The inductances and capacitances of the xtal are also used to close the transformer loop.

Yes those values are (very) critical for operation.

[Bricolo]

OK, so values are critical.

Elso said that ceramic NPO are the caps to use.
Is THIS also critical?

I can find NPOs here, but with lower values. And The required values, but not in NPO...

[Lars Clausen]

NPO's have a good low temperature coecifient, and are good for this application. Other ceramic caps may also work, but give a poorer temperature stability. Not nessescarily a problem.

You can also get better types: MICA caps are even better than NPO's, if you can get them.

C0G's are also good, same as NPO's or slightly better.

[Bricolo]

Is NPO meaning something like "low temp coef" ?

I agree that the tempco won't be a big issue here, I think the temp won't vary that much inside the player.

By mica, are you meaning silver mica?

And is C0G ceramic too?