32kHz Crystal oscillator with 74HC4060 - doesn't oscillate! (Now does!)

[Yansi]

Hello!

as I have mentioned in another thread about (still and again just other) FPGA minimalistic board, I like also to design small PLD kits. One of my boards includes Altera's EPM3064, a small 44pin CPLD. As a clock source, I decided to use a 32.768kHz crystal to provide both 2048Hz and 2Hz clocks, by using a "classic 4060" chip. But it doesn't work, I don't know why. That's the only thing preventing me from publishing the design for anyone interested.

Can you please help me to solve the mystery, why it doesn't oscillate?

As the whole board is powered from 3V3 supply, I have decided to use 74HC4060 version of the IC. The schematic diagram is below.

I have tried 7pF and 12pF Cload crystals, tried changing the loading caps to whatever crazy values up some hundred puff, tried changing the feedback resistor from 100k up to 1M. (The series resistor R41 is now shorted) Nothing helped. And yea, the IC is not blown. Sticking a finger on its input makes it count.

Do you have any ideas, what I have done wrong or how to design it properly, so it will oscillate?

Btw, have also checked if there is enough gain margin, and as I remember there was a great margin.

Thank you for help!

Yan

[PA0PBZ]

Broken crystal?

[Yansi]

Tried two new ones (7pf and 12pF) and a few random junk ones in the THT (metal tube) package. None worked.

[matseng]

Looking at some random 32K oscillators with the HC4060 it seems like the series resistor usually ranges from 50K up to 330K, and the parallel resistor us usually between 5M and 15M.

[station240]

After poking google I found this PDF
http://www.geocities.ws/k7hkl_arv/Cheap_Watch_Crystal_Oscillator.pdf

[Yansi]

Hmm. Interesting idea. I have tried to reduce the value of the series resistor to increase the drive level and gain margin to try to force the oscillation. Maybe I should try the opposite way. Reduce the gain margin and it might magicaly work.

So let's give it a shot.

[Yansi]

Whatta?! The damned devil oscillates on first try. 4M7 feedback, 330k damping resistor in series, 22p caps with a 12.5p Cload crystal.

That was solved fast. Doh... stupid me.

Thank you, as always :-)

Yan