Using a 32,768kHz crystal with 4060 frequency divider

[mduchalski]

Hello,

Recently I've been building a clock using 4000-series logic ICs. The counters/display drivers section of my circuit seems to work without a problem when using an Arduino to generate a 2Hz clock signal. However I am having trouble getting the oscillator part to work. The IC I've been using for this is CD4060BE 14-stage frequency divider with oscillator - http://www.ti.com/lit/ds/symlink/cd4060b.pdf. Now, figure 13. in the datasheet shows a "Typical crystal circuit" but doesn't give any specific component values. The datasheet of my oscillator - https://www.tme.eu/pl/Document/6efe9750ac67fc9b43f199686a1256c6/38_26_15_Series.pdf also only seems to have a generic diagram of a Pierce generator.
Not knowing much about oscillators, I've decided to go with one of the schematics from the web. I can't find my original source now, but the circuit itself can be seen in the attachment. So I've build it, replacing the 22pF cap from the original schematic with a 24pF one (closest one I had). The results, however, turned out to be poor. I don't have a scope, but instead of 2Hz I'd expect on Q14 I'm measuring between 3-8Hz with my multimeter, depending on for e.g. board's position on the table, distance to multimeter probe cable etc..
After this I've searched the internet for other circuits with this IC/32k oscillator and found that each one seems to use slightly different components, some even variable capacitors. I also looked for information on Pierce generators in general. This - http://www.crystek.com/documents/appnotes/Pierce-GateIntroduction.pdf - looks promising, but on the other hand circuit shown there lacks one of the resistors seen in 4060 datasheet (the current-limiting one). So in general I'm asking what may be wrong with my circuit, what components should I use or where should I look/what to Google to find out?

PS I've also attached photos of the board in case soldering or component positioning causes this problem.

[Zero999]

R15 is far too high. Are you sure you didn't mean 470R?

[mduchalski]

Alright, since posting I've tried to find the schematic I used and finally I found it - https://goo.gl/images/resCD2. Based on this R15 seems to be fine, and other schematics also have current limiting resistors in the hundreds of kiloohms. R16 is definitely wrong though - it should be 10M instead of 1M. This is also consistent with other information I found that for 32k crystal this should be 10~15M. I don't have any resistors above 1M so I can't test it immediately, but I'll edit this post if this'll resolve the problem. I'm sorry for my stupidity and wasting your time.

[Zero999]

Alright, since posting I've tried to find the schematic I used and finally I found it - https://goo.gl/images/resCD2. Based on this R15 seems to be fine, and other schematics also have current limiting resistors in the hundreds of kiloohms. R16 is definitely wrong though - it should be 10M instead of 1M. This is also consistent with other information I found that for 32k crystal this should be 10~15M. I don't have any resistors above 1M so I can't test it immediately, but I'll edit this post if this'll resolve the problem. I'm sorry for my stupidity and wasting your time.

On second thoughts you're right. The series resistor is high for such a low frequency crystal and yes, the feedback resistor should be 10M, not 1M.

If that doesn't work, try reducing the series resistor value, but to 330k first, then 220k, not 470R, which will overdrive the crystal.

[danadak]

This might help -


http://www.ti.com/lit/an/szza043/szza043.pdf


Regards, Dana.

[Benta]

R15 is about right, in some cases it can even go to 1 M.
The trick about 32 kHz crystals is not to overdrive them. They only need around 1 uW of power to oscillate, if you give them more they stop.
AT cut crystals are a different story.