900MHz beat freq. transmitter
in [Design]
|
Prev: DANGER DANGER THIRD DAY CPU FAN FAILURE DANGER DANGER. (AMD SimNow 4.5.2 has no CPU FAN emulator :( )
Next: dsPIC33f and SPI,SSEN too short
From: Mark Langley on 27 Mar 2010 03:24 Can anyone advise what type of equipment would be needed to create a _localized_ beat frequency (standing wave) on the order of a few KHz by engineering the phase relationship between two 900MHz PWM carriers? What degree of oscillator stability is implied? Mark Langley
From: Winston on 27 Mar 2010 18:44
On 3/27/2010 12:24 AM, Mark Langley wrote: > Can anyone advise what type of equipment would be needed to create a > _localized_ beat frequency (standing wave) on the order of a few KHz > by engineering the phase relationship between two 900MHz PWM carriers? > > What degree of oscillator stability is implied? That would depend on your requirement for beat frequency stability. The following is just an illustration and is a bad practical approach WRT stability (see last line): The Crystek CVCO55CL-0805-0900 will move it's output by 30 MHz for every 1.0 V difference on pad 2 (Voltage Control Pin). This would require you to hold the control voltage variation below 300 nV pk-pk to stay within say +- 6Hz stability of your beat frequency, provided that your reference oscillator has *no* instability or drift. Even then, you would probably require a crystal oven and super regulated supplies for both oscillators to get that level of performance. V(high)=2.50000000 F = 852.906000 MHz V(low)=2.4999997 F = 852.905990 MHz F(ref) = 852.900995 MHz F(highbeat) = 5005 Hz F(lowbeat) = 4995 Hz The actual stability would be worse because the part imposes (for example) a 1 KHz phase noise of about -79 dbc/Hz. --Winston -- Today's retailer is in an awkward position. He must assuage his visceral need to anger some of his clients while having to delight them sufficiently to guarantee repeat business. |