Radiated Emission from 125MHz oscillator

I have performed Radiated emission test for our boards.But it is crossing the limit at 125MHz and their harmonics like 250,375,500,625,750,875 and 1GHz.

Can someone suggest how the same can be solved.

Try some bypass capaciators located right at the oscillator. Try lowpass filtering the output if that is possible in your system. If you have broken ground planes under the oscillator, try soldering on jumpers across any cuts. Try soldering on a metal shield over the oscillator.

Is there any standard logic IC such as inverter on your board?
These ICs may radiate up to 4 GHz and higher. For example, NC7WZ04 Fairchild at 100 MHz input gives -70...80 dBc harmonics at 3 GHz. Try EMI/RFI filters, LPF at output and EMI shielding of board including output connector.

The main problem is that you are using and testing an OPEN circuit board with a high-frequency oscillator on it.
Whenever you have an oscillator on a board, then -typically all- such devices will radiate the first and higher harmonics around. You could use blocking capacitors on DC lines, etc. but the main problem is that you are sending a RF power along the PC board lines.
The best solution to the problem is to enclose your PCB into a shielded enclosure with good DC blocks on all other than RF lines. The RF signal output should be guided into a good coaxial connector for external use. If this connector is terminated with a 50 or 75 Ohm termination load, there should be no stray radiation from your enclosed board. In actual devices, some stray radiation may be detectable but within the EMC-standard limits.

what jiripolivka says plus you could try to reduce the level of feed back or gain of the circuit so the wave form is a better sinewave - try 22 ohms in series with emitter?
Frank

Are you talking about a "digital" (square wave) crystal oscillator? Or what is it? Would be worth to mention some more details than a single word in the question title.

Thanks for your input

Yes i am using a square wave crystal oscillator. Board is open, we are using a clock buffer as well after oscillator output as this clock is required at more than one place.

I guess, you can't actually tell, that the oscillator is the EMI source, it may be any device supplied by the 125 MHz clock? A strong even harmonics would suggest a different source than the oscillator or its direct output, assuming it's around 50% duty cycle.

Some points hav ealready been mentioned, particluarly need for continuous ground plane. Also supply bypassing and possibly filtering by ferrites for all digital devices is advisable. But you need to identify the source of radiation more exactly, e.g. by scanning the board with a small probe coil and a spectrum analyzer.