Hi friends,
I designed a DECT(GFSK modulation) unit.Testing PCBA's frequency is OK,but when i assemble it,turn the unit around testing frequency drift(radiation),i found the drift change from 3KHz to 28KHz with different angle,my specification is within 15KHz.What causes the phenomenon?
Thanks,
I found when I touched E-cap that is bypass capacitor of power trace for the microphone,the
frequency drift is more serious.I do know what happen and how to solve it?
Anyone can help me?
Is the PLL VCO frequency the same as the transmitter output frequency? If it is, it might be the transmitter energy getting back into the VCO and pulling the freuqency.
Completely agree with toonafishy. The "drift should then be reduced when turning off modulation. Other options would be signal pickup on VCO tuning line, insufficient shielding, ...
Looks that you have a decoupling and/or a layout problem.
Check the power supply decoupling capacitors around the PLL (to all the elements of the PLL including VCO) and also check for ground layout anomalies, as using the same ground for PLL and for microphone preamplifier.
You have not explained your circuit. If it is a PLL trying to lock the MODULATED output, then you have to ask a little about the nature of the modulation. If you are modulating with more 1's than 0's, your modulation average frequency will be off-center, and your PLL will correct that so that the "average" frequency is now locked. So the frequency for 1's will be nearly at channel center, and the frequency of 0's will be way out of band.
Is the PLL VCO frequency the same as the transmitter output frequency? If it is, it might be the transmitter energy getting back into the VCO and pulling the freuqency.
Thanks for all of kindness.The PLL VCO frequency is not the same.The fully integrated VCO operates at 4/3 of the desired LO frequency.Attached shematics.