Assuming that you're talking kHz all the way here, if you need more precision you either need to roll your own oscillator from a precision comparator and precision resistors, or you need to consider using a 4060 and an appropriate crystal, or some other crystal/counter scheme.
Get away from the two-R configuration and use only a single R from output back to the capacitor... and use a potentiometer as part of the R, to set F accurately.
...Jim Thompson
Go back and decide between KHz and MHz. What values of timing components are you using?
Hello,
I am having some problem with the 555 type oscillator. I am testing various brands and qualities of 555s and each one produces a different frequency for the same configuration of resistors (Ra & Rb) and capacitor (C). Even between 2 ics of the same brand (and batch) the frequency is different. In detail for a theoretical frequency of 72KHz, i get experimental values of 54MHz for "no-name" TTL-555s, 56.6MHz for "no-name" LinCMOS-555s and 67.6MHz for Philips TTL-555s. As you can see the difference is from around 20MHz downto around 4KHz. To sum up, is this normal or is there a problem with my circuit? Does this mean the precision of an unreliable "no-name" 555 (that produces inconsistent frequency) will degrade in the future due to wear out from usage or to increased temperature? Is there an other economical and fairly common option to produce stable and theoretically predictable digital clock at 76KHz?
thanks in advance @@
BUT! It will stay on frequency one set with a pot.
...Jim Thompson
Another batch of 555 frequency mysteries?
I was hoping to see your feedback about the very similar query you raised 5 days ago, subject '555 oscillator queer behavour'. For example, so far you haven't shown us your circuit, or provided details of what frequency measuring approach you use.
Sorry my mistake, all frequencies are KHz.
--- Not necessarily, since the timing components will have their own tempcos which _aren't_ ratiometric.
Not only that, if the timing cap is electrolytic its leakage current will also vary with temperature, throwing the frequency off.
For the OP's application, at the frequency he's working he probably won't be using an electrolytic, but you never know...
-- John Fields
The previous matter (problem) is still open. I was not able to reach any theoretical frequency value (with the common 555 circuit of 2 resistor and 1 capacitor) but after trying several "brands" of 555s i came very close to the desirable 76.8KHz (in particular 8KHz less). So considering i was in a dead end i decided to use the most "reliable" brand of 555 and to cook the resistor values so the 555 produces 76.8KHz. Now about the measuring instrument, i used a frequency- meter (Kikusui electronics: Multi-Counter, model 255) which i found at the univercity and know nothing else about. I tested its precision with a known clock and found it OK. The following small pdf file contains the 555 configuration:
Yes. That happens.
The answer is on the data sheet IIRC. Yup.
Check 'initial error' on page 6. Can be 3% ( presumably +/- ) making 6% difference between 2 samples possible. That's for a monostable too. You'll be using it as an astable and the numbers look worse ( unspecified max ). Reading between the lines, assume it'll be no better than 12-14% worst case.
Don't forget the tolerance on your Rs and Cs either !
Graham
Until the temperature changes.
Back in 1975 my boss got interested in cheap oscillators, and I looked at a bunch of RC circuits - the 555 got discarded very early. An emitter-coupled multivibrator was a lot better, but even that wasn't good enough - probably due to the temperature sensitivity of the Early effect.
Bob Widlar's LM322 (of blessed memory) got rid of the worst problems of the 555 by using an internal regulator to stabilise the voltage across the oscillator circuit, but it still wasn't as good as the emitter-coupled multivibrator (which probably came from Peter Baxandall).
------------- Bill Sloman, Nijmegen
Until the temperature changes.
Back in 1975 my boss got interested in cheap oscillators, and I looked at a bunch of RC circuits - the 555 got discarded very early. An emitter-coupled multivibrator was a lot better, but even that wasn't good enough - probably due to the temperature sensitivity of the Early effect.
Bob Widlar's LM322 (of blessed memory) got rid of the worst problems of the 555 by using an internal regulator to stabilise the voltage across the oscillator circuit, but it still wasn't as good as the emitter-coupled multivibrator (which probably came from Peter Baxandall).
------------- Bill Sloman, Nijmegen
I make a point of designing out presets whenever practical ! *Someone* is bound to be unable to resist fiddling !
Graham :-)
You got 56.6MHz from a 555 *holy cow*.
Before anyone else says it "Use a PIC" They have a built in oscillator and can give you the needed frequency on an IO pin.
Now that that's out of the way:
How stable? How predictable? How accurate?
76*32 = 2.432MHz. You can get 2.457 crystals off the shelf. The CD4060 or 74HC4060 would do nicely.True, unless you use a bad ceramic capacitor. Use an NPO or plastic capacitor.
In article , Pooh Bear wrote: [...]
... or forget to fiddle on the production line.
I'm glad to see you're aware of that one too !
I've seen presets set in *very* odd positions occasionally. I have next to no time for them. If you can't design a tolerant circuit you should learn how to !
Hilarious example. A design I reviewed had a 5.1V zener reference diode. Followed by a preset to trim out its tolerance. Did no-one consider using a bandgap reference ? Probably cheaper in total cost too !
Graham
Have something to add? Share your thoughts — no account required.
Ask the community — no account required