Temperamental crystal circuit

[ahydra]

Hi all,

No doubt some of you will have been following the saga of my PIC-driven clock project. Today it has taken a weird turn as the crystal circuitry has started to misbehave.

The schematic:

  |-<32768>-----|
  |             |
  --[10M]-[220k]--> out     Legend: <32768> = crystal, [ ] = resistor, ( ) = capacitor
  |      |      |                All junctions are connections
  --|>o--|      |
  |             |
 (33)          (33)
  |             |
  0V            0V

(|>o = NOT gate = 1/6 of a 4069)

Here's the puzzle. My circuit is powered by a 4.5V battery supply which is enough to power the PIC/display but doesn't seem to get the crystal ticking. I previously powered the 4069 with 9V, so I tried that and it works. It also continues to work if, after starting off at 9V you take the wire out and connect it to the 4.5V supply.

If you reverse that (start with the 4.5V batteries then connect it to the 9V supply) the crystal does NOT work. So,

1) Why this weird one-way phenomenon? I checked after disconnecting the 4.5V supply and the voltage was at zero everywhere, so I don't think there was any charge left in the capacitors or anything (they're tiny anyway).

2) Is there some simple fix e.g. reducing the values on the resistors so that it works at 4.5V?

3) I also have a 2MHz crystal - do I need to make any changes to the above circuit to get this working. (not so urgent)

Thanks,

ahydra

 

[Roff]

I don't know if this will make any difference, but you should be taking your output from the 4069 output.

 

[eblc1388]

As per Roff's suggestion but I would go a step further by adding another 4069 to buffer the oscillator output and then use it, provided that there are "free" 4069 sitting there.

 

[ahydra]

Neither idea worked unfortunately. I think the problem was either I had the circuit set up wrong or simply FM (F***ing Magic - i.e. it just randomly disappeared), as today I rebuilt it and it works, even using the same set of batteries as before.

Do either of you/does anyone else know what component values I should use for a 2MHz crystal instead of the 32768Hz? The load capacitance is 32pf, and from websites I have found the capacitors should therefore be about 54pf (I'm getting some 56pf's). But nobody seems to want to tell me the ideal values of the resistors.

ahydra