Hi, Steve. I'd guess you're using this ultra-simple digital square
wave oscillator (view in fixed font or M$ Notepad):
___ ___
| .---|___|--|___|----.
| | R A |
| | | |
| | |\ | |
| o----| >O----o------o-----o
| | |/
| |
| +|
| ---
| C ---
| |
| ===
| GND
(created by AACircuit v1.28.6 beta 04/19/05www.tech-chat.de)
This circuit will work with an inverter with schmitt trigger input
like the HC14. The input voltage will oscillate between the upper and
lower switching points, and the output will be a sqare wave with 50%
duty cycle.
You're experiencing one of the biggest limitations of this circuit.
The upper and lower switching points of the schmitt trigger are only
specified within a wide range, and they tend to drift with
temperature. That means your frequency will temperature drift far
more than you would expect from just component drift.
A second consideration is whether your supply is poorly regulated or
unregulated. The switching points of the schmitt triger can be seen
as a percentage of Vcc. Those percentages will change with varying
power supply voltage.
What to do? You haven't described much about what you're doing, and
you also didn't mention how much drift is acceptable.
Let's look first at what to do to to help your existing oscillator.
Start out by keeping total R (including your pot) in the range of 33K
to 1Meg. Less means higher output current and chip heating, and more
means leakage currents and other problems predominate. Second, don't
use the HC14 to drive a piezo or another high current load, for the
same reason. Replace your cap with a ceramic NPO or other cap with a
tolerance of 5% or less. Make sure the circuit board is clean of flux
or other contaminants that can cause leakage currents which can bollix
up the works. If your pot is remote, put a 100pF ceramic cap in
parallel with the timing cap to help squelch higher frequency noise
that might be affecting your timing. Make this oscillator at the pin
5-6 inverter, and tie the negative end of the cap directly to the GND
at pin 7. Try to get your supply as well-regulated as possible. And
most significantly, use a multi-turn pot if you can to set frequency
-- the single turn ones are pretty miserable about holding a precise
value sometimes, especially at the ends of the wiper.
If these don't do it, you might want to try a 555. An LM555 is
specified with temperature drift of .015% per degree C in astable
mode, and drifts .3% per volt change in power supply. You can be
pretty much guaranteed nearly all of the frequency drift will be due
to your cap and pot, if you use a regulated supply.
And if you're running on a supply of less than 5V or are running on
batteries, try the LMC555. It has similar specs, and can run on the
full range of HC supply voltages without problem.
http://cache.national.com/ds/LM/LM555.pdfhttp://cache.national.com/ds/LM/LMC555.pdf
Look in the data sheet for the typical astable circuit.
Hope this has been responsive. If not, please take a little time to
describe more fully what you're doing and what you need --
specifically:
* What's your supply voltage? Is it regulated?
* Do you just need a frequency, or do you need a 50% duty cycle square
wave?
* What kind of frequency stability do you really nned?
* Can you afford a separate chip for this?
Cheers
Chris