Information about offset PLL (OPLL)
It is difficult to find reference book.
Can someone kindly provde the data about OPLL (Offset PLL) ?
Thanks for your help.
Not sure of the terminology used, but offset PLL's I have seen have a standard PLL (VCO, Divider, Loop Filter, Reference clock), but at some place in the loop, like before or right after the frequency divider, there is a mixer and a bandpass filter where you can mix in a small frequency offset.
Say you have a 1 ghz vco, you are dividing by 100, and your phase detector frequency is 10 MHz. After the divider, if you mix in say 200 KHz to the 10 MHz divider output and filtered it so the low side went thru, then the PLL will lock up at a new frequency of 10.2 * 100 or 1020 MHz.
You do this offset lower frequency with a 2nd PLL.
You do this to keep the N ratio in the original PLL small to keep phase noise good.
offset PLL is to insert in the feedback path a mixer , this mixer will translate the VCO output to certainfreq say Fvco+F1 , and u have Fvco-F1 , these u wil select one of them , and this will feed to the divider , the ouput of divider
khouly
http://www.diva-portal.org/diva/getD...__fulltext.pdf
The paper said the OPLL can reduce noise in receiver band (Tx noise in Rx band).
Compare to conventaionl PLL, could someone kindly explain why ?
Thanks a lot.
That paper is interesting, but has limits. In that paper, the RF output is a direct connection to the phase locked VCO. As such, there is no local oscillator or image frequency signal that have to be filtered out. If you have a broadband receiver where the receive frequency is a short distance away from the transmit frequency, then you can not make a narrow enough bandpass filter for the receiver to block the transmit signal LO or Image signal (since your receive tuning bandwidth for a different t/r frequency pair might be in-band).
So with this scheme, you do not need to filter out-of-band transmit noise, as there is only the modulated VCO fundamental output to deal with.
UNFORTUNATELY, you are going to be limited to pretty low data rate and not too complex modulation schemes with this approach. That is because the modulated IF signal is passed through a phase detector, turned into an analog video voltage, loop filtered, and then used to drive the VCO tune port. IF there is a broadband modulation, it will not get thru the control loop filter (limited bandwidth to keep the loop stable). If it is a complex modulation (like 8 PSK), then the open loop phase shift of the control loop will tend to shift the higher frequency components of the modulation more than the lower frequency components, and you will smear the signal energy in the phase space. If it is a FSK type of modulation, you have to worry about the PLL unlocking, or at the very least operating both in both linear AND non-linear states during modulation transitions. Modulation like QAM would be impossible since the AM part would be stripped off. In fact, pretty much any non constant-envelope modulation may have some serious distortion that may not be repairable at the recieve end of the link.
So, if you have a large channel tuning bandwidth, a simple modulation like BPSK, and a low modulation data rate, this is a neat idea.
