series clock for cds - is it new?

[padamiecki]

Hello

Like Elso I am a chemist, but oposite I am weak in digital.

Please comment of a chematic of a clock that I found on some Polish site.
As I have read from a description, the oscillator is sonnected in series, not paralled like in most application.
The clock is very stable and its drift during a year was 0.
It could work as "slave" and might be steering from indepentent voltage sources or PLL.
To work as cd clock we have to ommit R5,6,7,8,C9,D1.
D1 should be replaced with a fixed capacitor.

Here is the link: http://stud.wsi.edu.pl/~sikrolb/sche...oscylator.html

[peranders]

This is a series resonance oscillator. I have used this circuit (very simlilar) for a digital clock (on the wall) and the frequency was 4.194 MHz.

If you are going to use such oscillator note that the crystal must be made for this operating mode because the stability gets really bad if you tune a parallel crystal to series resonance, ie 4 MHz parallel tuned to 4 MHz series.

[Bricolo]

Can you explain why stability goes bad when using a parallel crystal in series mode?

I'm reading Hans Camenzind's book (http://www.designinganalogchips.com/) and at page 158 he mentios series and parallel resonnance modes for a quartz.
He didn't precise that crystals are made or either one or the other mode, just that both are 0.2% apart.

BTW, in our (unmodded) cd players, is the classical cmos inverter/crystal combo a series or a parallel oscillator?


Mr Camenzind also proposes an optimized cmos oscillator, with a series resistance (more details about this in his great (and free!) book) has anyone tried it?

[padamiecki]

Originally posted by peranders
This is a series resonance oscillator. I have used this circuit (very simlilar) for a digital clock (on the wall) and the frequency was 4.194 MHz.

If you are going to use such oscillator note that the crystal must be made for this operating mode because the stability gets really bad if you tune a parallel crystal to series resonance, ie 4 MHz parallel tuned to 4 MHz series.

This means that for my Philips CD I have to use 11,28....MHz quartz?

[padamiecki]

Originally posted by peranders
This is a series resonance oscillator. I have used this circuit (very simlilar) for a digital clock (on the wall) and the frequency was 4.194 MHz.

If you are going to use such oscillator note that the crystal must be made for this operating mode because the stability gets really bad if you tune a parallel crystal to series resonance, ie 4 MHz parallel tuned to 4 MHz series.

... and series method realy gives more stable frq than parallel?

[peranders]

Originally posted by padamiecki

... and series method realy gives more stable frq than parallel?

Some people claims that and the reason would be that the series resonance is sharper than the "anti resonance" when you are using the "L" part of the behavior.

[peranders]

Originally posted by Bricolo
Can you explain why stability goes bad when using a parallel crystal in series mode?

When you tune a crystal much the Q factor will be worse (less stable oscillations) and the tuning elements will have more influence of the oscillation.

[peranders]

Originally posted by Bricolo
BTW, in our (unmodded) cd players, is the classical cmos inverter/crystal combo a series or a parallel oscillator?

Most of the time the latter.

[Bricolo]

can someone show us a basic series clock and a basic parallel one? I'm not sure of the difference

[Elso Kwak]

Hi padamiecki,
The circuit pictured is from the Elektor magazine. It is a Buttler oscillator if I am not mistaken.. This type of oscilator has certain advantages at very high frequencies like 100MHz. I tried all kind of oscillator circuits and found the Colpits the best though the differences are not big.
I found that a FET gave better results but the circuit could be modfied so.

Per-Anders you are completely wrong. There is no fundamental difference between a series and parallel resonant crystal. The only difference is the way it is calibrated. A series resonance specified crystal will oscillate at a slightly different frequency in a parallel resonant circuit like my clock.
Looks like you never studied the papers you promised to send me a year ago but never did.......