Resources regarding Phase Locked Loops?

[Michael]

Hi - I'm going to need to build a PLL circuit that can compare two
signals in the MHz range and find the phase difference between the
two. One will be very nice and clean while the other will be (I
suspect) quite noisy.I have never, ever done anything with PLLs. My
coursework at my uni (I am a senior EE, graduating in May) has not
covered PLLs at all. I don't think the term has ever even been
mentioned. I know about them from outside research.

Anyways - can anybody point me towards good resources regarding
designing such a circuit?

Thanks!

-Michael

 

[Joerg]

Hi - I'm going to need to build a PLL circuit that can compare two
signals in the MHz range and find the phase difference between the
two. One will be very nice and clean while the other will be (I
suspect) quite noisy.I have never, ever done anything with PLLs. My
coursework at my uni (I am a senior EE, graduating in May) has not
covered PLLs at all. I don't think the term has ever even been
mentioned. I know about them from outside research.

Has never been mentioned? Ouch. What university was that?

Anyways - can anybody point me towards good resources regarding
designing such a circuit?

If you've never done a PLL I suggest to get a copy of the ARRL Handbook.
Mine are older but PLL is handled in there. If this is for a mass
product and it's on a tight schedule get consulting help.

I think Analog Devices has some good app notes about the topic as well.
So did Philips but then they messed up their web site.

 

[DaveM]

Hi - I'm going to need to build a PLL circuit that can compare two
signals in the MHz range and find the phase difference between the
two. One will be very nice and clean while the other will be (I
suspect) quite noisy.I have never, ever done anything with PLLs. My
coursework at my uni (I am a senior EE, graduating in May) has not
covered PLLs at all. I don't think the term has ever even been
mentioned. I know about them from outside research.

Anyways - can anybody point me towards good resources regarding
designing such a circuit?

Thanks!

-Michael

Phase-Locked Loops, 5th Edition
Roland E. Best
McGraw-Hill, June 2003, ISBN 0071412018

Cheers!!
--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in the
address)

Life is like a roll of toilet paper; the closer to the end, the faster it goes.

 

[Michael]

Has never been mentioned? Ouch. What university was that?


If you've never done a PLL I suggest to get a copy of the ARRL Handbook.
Mine are older but PLL is handled in there. If this is for a mass
product and it's on a tight schedule get consulting help.

I think Analog Devices has some good app notes about the topic as well.
So did Philips but then they messed up their web site.

It's not that PLLs aren't covered at all - they're covered in some
classes, just none that I have taken. I took a pretty specialized
courseload, and PLLs are far from my specialty. However, it looks as
if I'm going to have to learn the buggers.

It's not for a mass product (at least not on a strict timeline -
eventually I hope to bring this device to market)

I'll take a look at the ARRL handbook. I see it mentioned enough that
I should check it out anyways.

-Michael

 

[martin griffith]

Has never been mentioned? Ouch. What university was that?



If you've never done a PLL I suggest to get a copy of the ARRL Handbook.
Mine are older but PLL is handled in there. If this is for a mass
product and it's on a tight schedule get consulting help.

I think Analog Devices has some good app notes about the topic as well.
So did Philips but then they messed up their web site.

http://www.standardics.nxp.com/products/hc/pdf/74hc7046a.pdf
http://www.standardics.nxp.com/support/pll/pll.zip


martin

 

[Joerg]

http://www.standardics.nxp.com/products/hc/pdf/74hc7046a.pdf
http://www.standardics.nxp.com/support/pll/pll.zip

That's the old 4046 style stuff. I thought Michael meant some "real" PLL
apps. Philips used to have nice app notes about that as well. Maybe they
still are somewhere on their server. It's just that it has been becoming
so fluff-laden and freaking slow. I understand that it's faster from
Europe but that doesn't help us much here in the Wild West.

 

[martin griffith]

That's the old 4046 style stuff. I thought Michael meant some "real" PLL
apps. Philips used to have nice app notes about that as well. Maybe they
still are somewhere on their server. It's just that it has been becoming
so fluff-laden and freaking slow. I understand that it's faster from
Europe but that doesn't help us much here in the Wild West.

I know its old, but the 7046 goes upto 20ish megs, and the basics are
covered in the data sheet. But Micheal didn't really say how little he
knows

I had a quick browse on th nxp site for the app notes, it seems you
have to email them for a copy, pathetic.

I'll try to find a rather good (IMHO) philips PLL appnote on my
backup drives/CD, and stuffit somewhare accesable


martin

 

[Joerg]

I know its old, but the 7046 goes upto 20ish megs, and the basics are
covered in the data sheet. But Micheal didn't really say how little he
knows

IIRC it's more like 17-ish. And probably only under the full "Princess
on a Pea" treatment, and when Mars is in the correct constellation and
no black cat has crossed the road from left to right.

I had a quick browse on th nxp site for the app notes, it seems you
have to email them for a copy, pathetic.

IMHO that whole web site has become pathetic, almost to the point of
being useless. And then some day they'll wonder why the sales numbers
don't come in as expected.

 

[[email protected]]

4046 or its superfast 74HC4046 or 77046 sisters are a good start for a
begining PLL person, but are junk compared to whats out there now.
However for about 3$ in parts plus a power supply and voltmeter or
perferably a oscilloscope, it will get you started learning and may
actually solve your problem, although a difference counter seems much
more like a solution for you

if you want to start with the 4046 as a learning experience, harry
Lythall's pages are great:

http://www.cqham.ru/projects/cmos_20rf_20synthesizer/cmos_20rf_20synthesizer.htm

http://web.telia.com/~u85920178/use/synth-00.htm

or the tutorials here:

http://web.telia.com/~u85920178/

click on "projects" then click on "synths"

Steve Roberts

 

[martin griffith]

IIRC it's more like 17-ish. And probably only under the full "Princess
on a Pea" treatment, and when Mars is in the correct constellation and
no black cat has crossed the road from left to right.

I always found the Lagrange Point pretty good for PLLs

IMHO that whole web site has become pathetic, almost to the point of
being useless. And then some day they'll wonder why the sales numbers
don't come in as expected.

martin

 

[Joerg]

I always found the Lagrange Point pretty good for PLLs

Isn't that where a big enemy spacecraft was hiding in Star Trek?

Anyhow, I prefer the liquid versions:
http://en.wikipedia.org/wiki/Image:Lagrange2000.jpg

 

[RST Engineering \(jw\)]

My God. My community college engineering TECHNOLOGY students get a full
week of PLL in the second semester of their FRESHMAN year. I'd like to
know, just to avoid, what university are we talking about?

Jim



I have never, ever done anything with PLLs. My

 

[Michael]

My God. My community college engineering TECHNOLOGY students get a full
week of PLL in the second semester of their FRESHMAN year. I'd like to
know, just to avoid, what university are we talking about?

Jim

You have got to be kidding me. Why post such nonsense? Is this some
sort of way to fluff yourself up? An attempt to put others down?
Perhaps a way of compensating for an insecurity about a shoddy
education, or a lack thereof? It's pitiful, sad, and a failure on all
counts.

To answer your question - I attend the 3rd ranked university for EE in
the USA.

I would suggest that your CC is teaching it too early, as there is no
way in hell that they have a strong enough background to understand
how it works at that point. Just having quickly glanced at the math
behind PLLs I can say that. Learning implementation should follow
understanding, in my opinion. I suspect those running your CC feel
differently.

-Michael

 

[joseph2k]

You have got to be kidding me. Why post such nonsense? Is this some
sort of way to fluff yourself up? An attempt to put others down?
Perhaps a way of compensating for an insecurity about a shoddy
education, or a lack thereof? It's pitiful, sad, and a failure on all
counts.

To answer your question - I attend the 3rd ranked university for EE in
the USA.

I would suggest that your CC is teaching it too early, as there is no
way in hell that they have a strong enough background to understand
how it works at that point. Just having quickly glanced at the math
behind PLLs I can say that. Learning implementation should follow
understanding, in my opinion. I suspect those running your CC feel
differently.

-Michael

Not a lot of ever EE's get to design PLLs. That should certainly make an
elective of the material. ETs often have to troubleshoot PLLs is existing
equipment, that is the difference engineers and technologists. To clarify
it is the difference between having to have the background to analyse it
for design, and having to have a basic understanding of typical industrial
applications and their various uses. Or quickly, between designing the
chip and using the chip.

 

[Winfield]

Hi - I'm going to need to build a PLL circuit that can compare
two signals in the MHz range and find the phase difference between
the two. One will be very nice and clean while the other will be
(I suspect) quite noisy.

Not sure what you're trying to accomplish, but I'd be inclined
to use a mixer, followed by a baseband filter, digitizing and
appropriate signal processing. An Analog Devices AD734 linear
multiplier would make a nice mixer.

 

[rebel]

Not sure what you're trying to accomplish, but I'd be inclined
to use a mixer, followed by a baseband filter, digitizing and
appropriate signal processing. An Analog Devices AD734 linear
multiplier would make a nice mixer.

If the two signals are locked (in a PLL ...) then mixing would be a fairly
pointless exercise. The O/P is after phase difference.

 

[Winfield]

If the two signals are locked (in a PLL ...) then mixing would be
a fairly pointless exercise. The O/P is after phase difference.

In a filtered DC mixer (multiplier) output the phase difference
shows up as a varying DC voltage which can be analyzed for phase
jitter and slower variations with time. I don't see the role of
a PLL, per se. PLLs are used to phase-lock an output oscillator
to an input signal. How does a second signal fit in, two PLLs?
Or are you only thinking of the PLL's phase detector? Remember,
this is after all just a mixer! But often, like in a 4046, it's
not a very good one, not as good as an accurate linear multiplier.

 

[Robert Latest]

a PLL, per se. PLLs are used to phase-lock an output oscillator
to an input signal. How does a second signal fit in, two PLLs?

Since the OP wants to find the phase difference of two input signals (one
clean, one noisy) maybe two PLLs wouldn't be such a bad idea. He could then
trivially mux the two oscillator outputs, both of which would be clean, of
equal amplitude and 50% duty cycle.

But as I've never done this I may be way off track here.

robert

 

[Fred Bartoli]

Robert Latest a écrit :

Since the OP wants to find the phase difference of two input signals (one
clean, one noisy) maybe two PLLs wouldn't be such a bad idea. He could then
trivially mux the two oscillator outputs, both of which would be clean, of
equal amplitude and 50% duty cycle.

But as I've never done this I may be way off track here.

Cleaning a signal with a PLL is just getting the signal through a narrow
BPF (and adding some noise and phase error too).
Mixing two PLL'ed signals the translates your BPF center frequency to
zero, transforming it to an LPF.

Multiplying the signal first and LPFing it after is just the reversing
the operations (first translate your frequency, then LPF it).

They are both mathematically equivalent, but I'd try to avoid the PLL
noise.
If the frequency is low enough I'd go for the multiplier first and LPF
later, and otherwise probably for a DBM and LPF.

 

[Robert Latest]

Cleaning a signal with a PLL is just getting the signal through a narrow
BPF (and adding some noise and phase error too).
Mixing two PLL'ed signals the translates your BPF center frequency to
zero, transforming it to an LPF.

Multiplying the signal first and LPFing it after is just the reversing
the operations (first translate your frequency, then LPF it).

They are both mathematically equivalent, but I'd try to avoid the PLL
noise.

Yeah, sounds right. Like I said, I'm out of my depth here.

robert