From: "Christopher R. Carlen"
Subject: PLL phase detector in CPLD
Date: Thu, 05 Dec 2002 16:28:05 -0800
Organization: Sandia National Laboratories, Albuquerque, NM USA
NNTP-Posting-Date: Thu, 5 Dec 2002 23:26:39 +0000 (UTC)
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.1) Gecko/20020826
X-Accept-Language: en-us, en
I am planning a design/PCB that will implement two PLLs for motor
control among other things. Other logic needs of the system lead me to
choose to put a low to mid-sized CPLD on the board. I am considering
putting the two phase detectors in the CPLD. The phase detectors that
I'm considering are 3-state PDs, using a pair of FFs and an AND gate.
The output is differential so will need to go to an op-amp to condition
before the loop filter, obviously.
What I'm concerned about is that the logic levels output by the PD bear
upon the PD output voltage. If the CPLD is doing other things, then
there may be noise in the power supply, low as well as high frequency,
that may affect the quality of the PLL.
Alternatively, I could put the phase detectors into discrete logic,
powered by their own dedicated regulator. Then I could be sure the PLL
won't get bumped around by power supply noise. But I don't like this
because I want the flexibility of changing the logic at any time, even
for the PDs, because it is possible I might want to experiment with
higher N-state PDs. I have been burned recently by the consequences of
designing discrete logic, after I made a fairly complicated thingy about
a year ago with HC, but now they tell me they need it to do something
The design and PCBs I made so far for that thing, are now junk because
at the time I didn't know CPLDs, and designed hard-wired logic. Never
again! From now on if it has any possibility of having a reason to be
modified or enhanced in the future, I'm building with all PLDs.
I suppose the best thing to do in this case is just to use two CPLDs,
one little one like a Xilinx XCR3032 for the PDs, powered by a dedicated
regulator, and another one a little bigger for the general logic.
Any better ideas?
Thanks for comments.
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA