Re: Crystal Oscillator problem
- From: Phil Hobbs <hobbs@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>
- Date: Mon, 10 Sep 2007 13:32:22 GMT
Mig wrote:
I encountered with very strange problem with a ceramic crystal
oscillator.
It took a long time to recreate the problem.
When a power supply voltage is turned on first time there is no output
frequency. When the power supply is turned off and on again the
frequency appears.
You can see it on the video here.
http://www.youtube.com/watch?v=9mWOWL4LGsU
Any suggestions?
The vendor doesn't know why it happens. :(
Startup problems are as common as dirt.
Most oscillators work in a large-signal regime where the active devices are nonlinear. Thus the average gain, impedance level, and feedback are all different when the oscillator is running, compared with the same circuit in a quiescent state.
The feedback loop may well be stable at small-signal conditions and only oscillatory at large signal conditions. If the power supply transient is quicker than the bias time constant of the oscillator, it'll find itself in a large-signal condition at power-up, which will get the oscillator going and mask the startup problem. It's really important to test oscillators under a wide variety of turn-on conditions, including very slow ramps.
It's also quite possible for oscillators to have too much feedback, in which case you get all sorts of distortion and self-modulation rather than a nice CW sine wave. ALC oscillators adjust their loop gain based on oscillation amplitude, and thus tend to avoid both kinds of problem (as well as being much quieter).
I gather that your power supply is producing different turn-on transients, depending on how long it has been switched off. From the symptoms, it appears that when it's been off long enough for the filter caps to have discharged, the turn-on is slower than when you turn it off and then on again, and the difference is enough to expose the startup problem.
Cheers,
Phil Hobbs
.
- References:
- Crystal Oscillator problem
- From: Mig
- Crystal Oscillator problem
- Prev by Date: Re: What's the Toughest Branch in Electronics?
- Next by Date: Re: getting started on C programming on PIC16F628A, PIC16F690, OR PIC18F2620
- Previous by thread: Re: Crystal Oscillator problem
- Next by thread: Biomedical-Bioinformatics Engineering
- Index(es):
Relevant Pages
|
