Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Analog PLL design, mixer as phase detector

Status
Not open for further replies.

Antenna94

Junior Member level 3
Joined
Apr 13, 2016
Messages
30
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
312
Hi!

I'm trying to design an analog PLL for fun with about 25 or 50MHz frequency. I was planning to use a Gilbert-cell mixer as a phase detector, but I have a question. A transistor mixer needs a bias circuit, and that should be decoupled from the other parts of the circuit. But by using capacitors, the phase detector output will not have a DC component. I would think that a DC output is needed to maintain phase lock. Am I right? Or does the jitter because of temperature and stuff of the VCO make it okay?

Thanks in advance!
 

That depends on the output you need to obtain and the signal you want to "phase-lock" with.
 

Phase detector output will be DC coupled in all usual PLL designs.
 

FvM but then how can you ensure that the output is DC coupled but the bias voltage doesn't get out? Or we cannot use a transistor mixer as a phase detector, only diodes?
 

Mixer is a linear PD and its output will be DC term+ 2X harmonics. The 2X harmonics will be filtered out by Mixer intrinsic output capacitor, while the DC term is proportional to the input phase error (possibly within +-90 degree). So no need to use the capacitor to block the output of mixer. You can sweep the charge pump output current with the phase error (the gradient should be Kpd), if there is a static current even when the phase error is 0, then resize the CP transistors.
 

Maybe my question isn't clear, or I just don't see the answer in the replies. I am asking if it is possible to use a transistor mixer as a phase detector, no charge pump. I understand that there is "no need to use the capacitor to block the output of the mixer", but then the DC bias of the transistors get out to the output.
 

It looks like you didn't yet start to analyze the behavior of your VCO design, otherwise you should be able to answer the question yourself.

No matter if you use a charge pump, or other circuit elements, you set up a control loop. The circuit between the phase detector output and VCO input defines a transfer function. A "wire" would be also a kind of transfer function, if it's suitable depends.

As already mentioned, you need a non-zero transfer function gain at DC to achieve PLL lock. The said "DC-bias" problem might require a level shifter. You also need a low-pass filter to convert the pulsating mixer output into a steady VCO tuning voltage. The filter characteristic should be selected so that it achieves sufficient loop phase margin in combination with the integrating PLL characteristic.
 
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top