The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> The accuracy of oscillator's frequency
https://designers-guide.org/forum/YaBB.pl?num=1157956352

Message started by caosl_zju on Sep 10^th, 2006, 11:32pm

Title: The accuracy of oscillator's frequency
Post by caosl_zju on Sep 10^th, 2006, 11:32pm

hi,all
I need a 100Khz square wave oscillator.The measure of formerly designed one has a large departure to spec.It's architecture is composed of a current reference,a mosfet capacitor,a comparator.The deviation may be caused by current or mosfet capacitor.Could anyone tell me a better method to improve the oscillation frequency's accuracy? Thanks very very much.

Title: Re: The accuracy of oscillator's frequency
Post by loose-electron on Sep 11^th, 2006, 12:46am

Two immediate things come to mind here:

Capacitance accuracy - MOS Caps are nonlinear, suggest you go to a MIM type that exhibits more linearity. Sometimes this is also poorly modeled.

Current Accuracy - How is your current being defined? Unless there is an external current reference, or some trim/adjustment, most internally defined current sources have an absolute value error that is pretty large.

Two good places to start. After that, you need to start looking at the min-max accuracy/tolerances of all the parts and pieces.

Title: Re: The accuracy of oscillator's frequency
Post by caosl_zju on Sep 11^th, 2006, 1:44am

hi,loose-electron.
Thanks a lot .It looks as if that the current is a troublesome problem except for trimming.but the cost will be increase.

Title: Re: The accuracy of oscillator's frequency
Post by loose-electron on Sep 11^th, 2006, 10:17pm

If you want an accurate current (without trim) use an external reference resistor.

Bandgap drives a V to I converter, which uses an external reference resistance.

There are gobs of chips out there that have been doing this for years.