I used a cheap cheap clock

[fmak]

Originally posted by Elso Kwak

-----------------------------------------------------------------------------------



Jung circuit not suitable for clock supply; been there done that.
You like to screw quoting systems fmak, jeeez!

---------------------------------------------------------------------------------

Your circuit might not have worked. My ALW versions power clock circuits beautifully. Dead stable and 10 uV noise/1 MHz. Also monitored on 400 MHz scope.

What do you mean by 'You like to screw quoting systems fmak' ???

Adding capacitance to XO before HC-- can help actually. Basically, matching is important.

[davidsrsb]

No capacitance on the 1st reg O/P -> 2nd I/P Amazing that it doesn't oscillate. Decoupling the adj pin to ground probaly will then.

The thick wires are so stiff that they are the real suspension, not the rubber bands.

[Elso Kwak]

I tried the ALW version....

You need a low noise RF-regulator for clock circuits [no more hints]

[Elso Kwak]

Originally posted by davidsrsb
No capacitance on the 1st reg O/P -> 2nd I/P Amazing that it doesn't oscillate. Decoupling the adj pin to ground probaly will then.

The thick wires are so stiff that they are the real suspension, not the rubber bands.

Yes you need very flexible wires indeed. Yes the LM317 not applied as in datasheet!

[pwillard]

More peanut gallery comments...

National semi *recommends* using a decoupling cap (.1uF) on the LM317 input pin. Why take chances... there should be one. I also feel that a 470uf main filter cap is under sized. I'd want to see at least twice that amount IE; 1000uF.

The LM317 is a 1.5AMP device. Under worst case conditions, this circuit draws 100ma, which is right at the hairy edge of what single 78L05 can deliver. Clearly, to be cautious, a smarter choice would have been a single LDO LT1086-5 device from Linear Technology. You get the benefit of Low Drop Out protection and 5 volt regulation and minimal parts count. The logic behind why two LM317's were used escapes me since attempting constant current regulation is overkill here.

The HEX inverter is almost entirely wasted. A buffer was created by using two inverter gates, basically keeping the Oscillator output in phase with inverter output... but why. No real need that I see... though it does look like they were going for a sort Schmitt-Trigger circuit with the inverter.

I would have used something like a buffer with open-drain output also known as 74AHC1G07 from Philips and would have the open drain buffer output tied to VCC though a 10K resistor so you get a TTL level HIGH when the gate is OFF. The 74ahc1g07 is a single gate, 6 pin, small footprint device. Ideal for use in a specialized circuit such as this.

But hey, I'm NOT a designer.

[fmak]

Originally posted by Elso Kwak
I tried the ALW version....

You need a low noise RF-regulator for clock circuits [no more hints]

-----------------------------------------------------------------------------------

May be your clock, but not others. I can measure no appreciable RF noise. What about all these clocks around that are powered by 7805s?

Strange coming from someone who likes supersonic spikes in non OS DACs!!

Why should I not use my own system of referral

[Panelhead]

Originally posted by pwillard
More peanut gallery comments...

The LM317 is a 1.5AMP device. Under worst case conditions, this circuit draws 100ma, which is right at the hairy edge of what single 78L05 can deliver. Clearly, to be cautious, a smarter choice would have been a single LDO LT1086-5 device from Linear Technology. You get the benefit of Low Drop Out protection and 5 volt regulation and minimal parts count. The logic behind why two LM317's were used escapes me since attempting constant current regulation is overkill here.

But hey, I'm NOT a designer.

PW,
I am really cheap, so this clock is interesting. The clock module looks like the Tent modules.
The power supply looks more like something desgned to power slowet devices. To impove on this, would using LT1086CT here instead of LM317 help any?

George

[Sunsun22]

Wow wow wow, this thread attracts so many responses. Thank you again for all the input.

Please bear in mind this is a cheap cheap kit but the fact is it improved the sound quality of my vintage Pioneer T-07A to great extend. I am using the T-07A as drive only and is connecting it to Dack 2.0 NOS DAC. Perhaps the age of my T-07 is a problem that a small modification generates so much improvement.

My last question is about the numbers of available clocks in the market that could cost 10x of this cheap cheap kit is numerous, can I expect to improve my CD by even greater extend by installing any of them in my aging T-07A? Or is it worthwhile to modify this cheap cheap clock to perform even better?

.

[fmak]

My last question is about the numbers of available clocks in the market that could cost 10x of this cheap cheap kit is numerous, can I expect to improve my CD by even greater extend by installing any of them in my aging T-07A? Or is it worthwhile to modify this cheap cheap clock to perform even better?

------------------------------------------------------------------------------------

If you really want to know, send it to jocko or JohnW (search diyhifi.org) for measurement. Many of the leading XO modules are made to order in China!! No one can tell you objectively without measurement.

The other way to do it is to use best practice in terms of PS and coupling, and pick the one that sounds best!!

[Ruach]

Anyone use this clock upgrade?