32.768kHz Pierce Oscillator (SN74LVC1GX04)

Community Projects
1

I’m trying to build a RTC clock driver for a CPLD board using a 32.768kHz tuning fork and a [Pierce Oscillator, but I’m having trouble getting the component values to work out right, and I’m currently not getting any oscillation out of it. I don’t have access to an oscilloscope right now, so I can’t see what’s going on on the clock signal line in detail, but it seems to be holding at a logic high, so to start out, I just want to figure out if I even chose reasonable values for the various components of the circuit, because after doing the math, several of the components I ended up with seem kind of extreme to me. I’ve looked at several references, but [this document has been the best at laying out the component selection. The inverter I’m using is the [TI SN74LVC1GX04, and the crystal is rated for a load capacitance of 6pF. According to the inverter’s datasheet, the input capacitance is 7pF max. So the external load capacitors are calculated by the formula:

         -(C1 + Cin)(C2 + Cout) -
Cload = | ---------------------- | + Cstray
         - C1 + Cin + C2 + Cout -

so for my 6pF crystal, and assuming a stray capacitance of 2pF, I get

       -(C1 + 7pF)(C2 + 7pF) -
6pF = | --------------------- | + 2pF
       - C1 + 7pF + C2 + 7pF -

with C1 = C2, that works out to exactly 1pF… which seems really low, compared to the ~20-30pF I typically see suggested.

Next up, RF is pretty straightforward. From everything I’m read, the value is not critical, and the document I linked suggests 10-15Mohm for 32.768kHz. 10Mohm is the biggest I have on hand, so I’m going with that.

Finally, RS is calculated by the formula:

     1
------------
2(pi)(f)(C2)

so

          1
--------------------
2(pi)(32.768kHz)(1pF)

which comes out to ~4.857Mohm… which seems awfully high, considering that earlier in that document, they talk about tuning fork crystals needing “large” values for RS >10K, so if 10K is “large” then 500x that is… well… REALLY large.

Am I just going crazy, or did I miss something in the math somewhere?](http://www.ti.com/lit/ds/sces581b/sces581b.pdf)](http://www.crystek.com/documents/appnotes/Pierce-GateIntroduction.pdf)](Pierce oscillator - Wikipedia)

2

At first glance, your Rs is way too high. Try in the order of 100 to 1k Ohm. I’ll take a closer look at your math when I get home.

3

rrpilot:
At first glance, your Rs is way too high. Try in the order of 100 to 1k Ohm. I’ll take a closer look at your math when I get home.

From the document I linked in my post (which is what I was using for my formulas):

Rs must be used with “Tuning-Fork” (watch) crystals. Tuning-Fork crystals have a maximum drive level of 1µW maximum. Without a large Rs (greater than 10k ohms), the inverter will physically damage the crystal!

I tried values as low as 10K, still no oscillation.
4

What is your power supply voltage?

What is the DC voltage your measuring at the output?

What is the AC voltage you’re measuring at the output?

5

Power supply is 3.3V, I’ll try to get the other readings when I get home tonight.

6

My multimeter’s battery died, so I didn’t get the chance to check it out until today, and whaddya know, it works just fine -_-.

The only thing I can think of is that maybe there was still some water under the chip from washing it off after soldering, and it finally dried after a few days. Sheesh…