Searching \ for 'PLL [Sot]' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/logic/dsps.htm?key=pll
Search entire site for: 'PLL [Sot]'.

Truncated match.
PICList Thread
'PLL [Sot]'
1999\08\09@202018 by Vincent Deno

flavicon
face
Hello again,

Was just wondering if any of you electronic gurus out there could help
with a PLL problem.

The problem is we need a PLL circuit which can vary freq. by about 1200
PPM with center frequency at 16 MHz.  Problem is most VCXOs are
manufactured to around 100 PPM or better.  A _poor_ crystall oscillator
(which allows a lot of freq swing) is still not sufficient, and is even
harder to control when it comes to introducing jitter.

Any ideas would be greatly appreciated.

Digital frequency sysnthesizers are being looked at, but we still need an
analog circuit which works.

Thanks,

Vincent Deno

1999\08\09@220129 by Dennis Plunkett

flavicon
face
At 20:19 9/08/99 -0400, you wrote:
{Quote hidden}

How many PPM do you need? adn whatamount of jitter? You will find reference
crystals with 10ppm at 10mHz very easy and cheep.


Dennis

1999\08\10@152915 by Vincent Deno

flavicon
face
> How many PPM do you need? adn whatamount of jitter? You will find reference
> crystals with 10ppm at 10mHz very easy and cheep.
>
>
> Dennis
>

We are in need of a reference crystall with a swing of about +-1200 PPM at
Fc of 16MHz and a couple other frequencie in that range.  Obviously, 10PPM
is not acceptable.  Thanks for the response nonetheless.

-Vincent Deno

1999\08\10@154638 by Wagner Lipnharski

picon face
> We are in need of a reference crystall with a swing of about +-1200 PPM at
> Fc of 16MHz and a couple other frequencie in that range.  Obviously, 10PPM
> is not acceptable.  Thanks for the response nonetheless.
>
> -Vincent Deno

So, what you want is a temperature sensor that changes 19200 Hz/¡C
running in a base frequency of 16MHz.  That margin of high ppm is also
directly related with (somehow) bad quality of components, and probably
you will not get a steady and linear response, with lots of differences
from one component to another.  Probably you should go to measure that
temperature using some other technique.
Wagner

1999\08\10@163712 by Matthew Fries

flavicon
face
> The whole point of a PLL is that the oscillator is free running, not a
> crystal one, this allows a large frequency swing. This signal is then
> divided down and compared to a reference crystal oscillator (also
> usually divided down!), this process produces an error signal, which is
> used to correct the free running oscillator which locks the oscillator
> solidly to the crystal reference. The output frequency can be changed by
> altering the division ratio in the divider chain, the actual resolution
> is dependent on how this is done - basically (and simplified) if your
> crystal reference is divided down to 1000Hz you will have 1000Hz
> resolution.

One thing I cannot seem to get my mind around is why... Why lock a an
oscillator's phase to a crystal reference. Why not just use the crystal
frequency and divide it as necessary? Is it the shape of the output wave
or something?

1999\08\10@193056 by Norman Gillaspie

flavicon
face
It it easy to provide various solutions. However he has not said
what he wants to do exactly or why. He just needs a crystal
oscillator or low phase noise source he can control with a voltage.

He might want to lock the crystal to a signal whose frequency
varies for some reason; and the signal he wants to use as a reference
is needs to either some multiple of this reference or that is phase
coherent with this reference signal.

This is a guess as he has not told us why he needs this reference.

Norman


PCS Engineering
Norman Gillaspie
325M Sharon Park Dr. #210
Menlo park, Ca. 94025
Tel 650-854-5263
Fax 650-854-5445
KF6WHG
Email spam_OUTnormanTakeThisOuTspampcseng.com


>

1999\08\10@234957 by David Bengtson

picon face
Do you need this to be continuously variable, or could you channelize
this? 600 ppm at 16 Mhz come out to 9600 Hz, so you need to hit about
+/- 10 KHz. DDS would make this pretty easy. Analog devices has some
chips that might work. Look for the 9851 on the AD web site.
Failing that, you could warp a very high frequency fundamental
oscillator and mix down. If you were at 160 Mhz, you would only have
to warp +/- 60 PPM, which is acheivable. Then you could mix down to 16
Mhz by using a 146 or 176 Mhz Local oscillator. More complicated this
way.


Good luck,

Dave Bengtson



On Mon, 9 Aug 1999 20:19:06 -0400, in  you wrote:

{Quote hidden}

1999\08\11@105726 by Scott Dattalo

face
flavicon
face
Some time ago Mike Keitz posted info about a PIC pll that was designed to
lock to HSYNC of an NTSC signal. His approach was to vary the pic's clock
frequency using the variable capacitance of a reversed biased diode. IIRC,
the diode provided some of the capacitance of a resonator circuit. I don't
see why this approach wouldn't work here as well. I'm not sure how your
feedback circuitry is implemented. In other words, does the signal you
generate get compared to something else to create feedback control (e.g.
like Mike's HSYNC) or is it generated as an absolute (e.g. something tells
your system to spit out a square wave with a xx kHz frequency). If it's
the former then you'll obviously need the conditioning circuitry to convey
the error to the pic.  If it's the latter then you'll need a reference
clock source to which you can compare your generated signal. This is
analogous to the DDS solutions others have discussed.

Perhaps Mike could explain the concept more clearly (or repost that old
message). But as I recall, he varied the control voltage on the diode by
creating a bit-banged pwm signal that was averaged with an RC circuit. The
RC circuit's time constant was much slower than the HSYNC frequency. Every
time an HSYNC edge came in, the pic would compare it to when it 'expected'
the HSYNC to occur. If the expected edge was too fast then the pic was
slowed down by making the PWM output low. If the expected edge was too
slow then the PMW output was made high. Pretty damn clever.

Scott

1999\08\11@115056 by Vincent Deno

flavicon
face
Yes, the freq will need to be continuously variable.  The DDS solution
will work for one design.  However, it will require horsepower from a proc
that is not there in a current design.  Therefore, it is necessary to find
another solution.  What is really needed is the equivalent of a 16MHz VCXO
with a pull of +-1000 PPM... while avoiding the $25+ each for such a
device.

Thus far, we've been able to get "acceptable" stability by putting an
inductor in series with the crystal to counteract the internal
capacitance.  But this requires further elemments for tuning (which is not
a problem), but still only gives a pull of around 500 PPM.

Any suggestions on how to increase the pull of crystal osc. woul be
appreciated.

> Do you need
this to be continuously variable, or could you channelize
{Quote hidden}

1999\08\11@115923 by Vincent Deno

flavicon
face
The currently implemented circuit is similar to what you describe.  The
problem s, the input signal is varying too much and is outside of the pull
of the crystal.  We're currently using some expensive varactors (~$5/each)
and a pretty good feedback circuit.  Problem is circuit was not designed
to pull the clock that far from its center frequency.

In case anyone is interested, this is for a audio/video application.
Problem originates in these cheap boxes which are coming out that pay very
little attention to their clock.  Therefore, we are forced to deal with
horrible digital timing--which is where the +-1000 PPM pull comes in.
-Vince

{Quote hidden}

1999\08\11@175558 by Nigel Goodwin

flavicon
picon face
In message <.....Pine.GSO.4.10.9908101529180.19279-100000KILLspamspam@spam@isis.visi.com>,> Matthew Fries <freezespamKILLspamVISI.COM> writes
>> The whole point of a PLL is that the oscillator is free running, not a
>> crystal one, this allows a large frequency swing. This signal is then
>> divided down and compared to a reference crystal oscillator (also
>> usually divided down!), this process produces an error signal, which is
>> used to correct the free running oscillator which locks the oscillator
>> solidly to the crystal reference. The output frequency can be changed by
>> altering the division ratio in the divider chain, the actual resolution
>> is dependent on how this is done - basically (and simplified) if your
>> crystal reference is divided down to 1000Hz you will have 1000Hz
>> resolution.
>
>One thing I cannot seem to get my mind around is why... Why lock a an
>oscillator's phase to a crystal reference. Why not just use the crystal
>frequency and divide it as necessary? Is it the shape of the output wave
>or something?

No, it's so you can get a much wider range of output frequencies, if you
divide down from a crystal you can only get a very small range. For
example, if you use a 1MHz crystal the highest frequencies you can
divide to are 500KHz, 333KHz, 250KHz as you can only divide by round
numbers. Using a PLL with the crystal divided down to 1KHz, you could
get 999KHz, 998KHz, 997KHz and so on, right down to 1KHz, if you need
greater resolution than that, divide the crystal frequency down further.
--

Nigel.

       /--------------------------------------------------------------\
       | Nigel Goodwin   | Internet : .....nigelgKILLspamspam.....lpilsley.demon.co.uk     |
       | Lower Pilsley   | Web Page : http://www.lpilsley.demon.co.uk |
       | Chesterfield    | Official site for Shin Ki and New Spirit   |
       | England         |                 Ju Jitsu                   |
       \--------------------------------------------------------------/

1999\08\11@192845 by Mike Keitz

picon face
On Wed, 11 Aug 1999 11:58:48 -0400 Vincent Deno <EraseMEdenovjspam_OUTspamTakeThisOuTEMAIL.UC.EDU>
writes:

>In case anyone is interested, this is for a audio/video application.
>Problem originates in these cheap boxes which are coming out that pay
>very
>little attention to their clock.  Therefore, we are forced to deal
>with
>horrible digital timing--which is where the +-1000 PPM pull comes in.

You've talked about "pull" a lot but that's only part of the picture.  I
can imagine at least 3 key specifications:

Pull-- If the control voltage is varied from minimum to maximum, how much
does the frequency change?

Accuracy-- If a particular control voltage is applied, how close is the
frequency to the expected value?

Stability-- With the control voltage held constant, how much does the
frequency change due to other factors?  Poor stability of course implies
poor accuracy.

If you need a circuit with large pull and high accuracy, it's going to be
expensive.  Most likely you don't, especially if a constant input signal
is present that your circuit can phase or frequency lock to.

An R-C oscillator can pull over a tremendous range.  Stability is
extremely low.  R-C oscillators phase locked to crystals seem to be in
all computers to generate the CPU and video clocks.

Rather large "pull" (at least tens of percent) can be acheived with an
L-C oscillator.  Since this is also a cheap circuit, the two out of 3 law
dictates that accuracy/stability will be poor.  It may be necessary to
adjust each unit at the factory to a frequency close enough that lock can
be achieved.  If the application requires less that extreme pull it is
much better than an R-C circuit.  I recommend using an IF transformer as
the L-C circuit since they have a capacitor temeperature compensated for
the inductor built in.

Next on the list of decreasing pull but increasing accuracy is a ceramic
resonator.  Most late-model TV sets use a ceramic resonator circuit for
horizontal and vertical timing.  It is locked to the incoming video's
sync.  The pull range is much larger than a crystal, probably enough for
1000 ppm, and the accuracy is not terrible.  However you will need to
work closely with the resonator manufacturer to make sure they will
continue to make units that work in your circuit.  Resonators designed
for "pull" circuits may be custom parts.

Finally there's the VCXO.  It has high accuracy and stability but very
limited pull.  Because of the linited pull, it's limited to situations
where the input frequency is already quite precise, such as a GPS
receiver or the color decoder of a TV.  In the former case it's stability
is a great advantage because the input signal may be lost for a while.
The receiver's local clock can stay close enough that the signal can be
re-synchronized quickly when it returns.

Without knowing about your application, I couldn't say if various
"hybrid" techniques could be used.  For example, an oscillator with poor
accuracy could be periodically re-calibrated to a fixed crystal
oscillator.  Or the pull of an oscillator could be increased by using
digital techniques to modify the output frequency.  A "pretty good
feedback circuit" can be very useful.  Feedback can make many
"impossible" things possible.




___________________________________________________________________
Get the Internet just the way you want it.
Free software, free e-mail, and free Internet access for a month!
Try Juno Web: dl.http://www.juno.com/dynoget/tagj.

1999\08\11@194352 by Norman Gillaspie

flavicon
face
FYI,

How about a VCO that has a center frequency of one of
the following center frequencies.

16.00
16.257
16.384
16.934

5 Volt Vcc
+- 2.5 volts tuning volatge equals +- 2500 PPM linear change in frequency
Temperature stability 0 to 50 degree C is 25PPM


I also have a single crystal design that has high pullabiliy but is not as
slick
as the above and has a lot more componenets.

Single qty price for the device is 25.00 Apx and samples maybe available.

I have been down this path before and have the experience to solve your
probelem
on a consulting basis.

So far Vince has not contacted me.



Norman



PCS Engineering
Norman Gillaspie
325M Sharon Park Dr. #210
Menlo park, Ca. 94025
Tel 650-854-5263
Fax 650-854-5445
KF6WHG
Email normanspamspam_OUTpcseng.com

1999\08\11@223821 by Dennis Plunkett

flavicon
face
At 16:48 11/08/99 -0700, you wrote:
>FYI,
>
>How about a VCO that has a center frequency of one of
>the following center frequencies.
>
>16.00
>16.257
>16.384
>16.934
>
>5 Volt Vcc
>+- 2.5 volts tuning volatge equals +- 2500 PPM linear change in frequency
>Temperature stability 0 to 50 degree C is 25PPM

That's a strainge range with only 5V. I would have thought that the tuning
range would have been only a few volts for simplistic sake. Yes you can use
a higer varactor voltage.

{Quote hidden}

Go to Varil, they have these in modules off the shelf and all you have to
do is add a PIC to set the frequency, and they should be somewhat less than
$25

Dennis

1999\08\12@121723 by Vincent Deno

flavicon
face
> You've talked about "pull" a lot but that's only part of the picture.  I
> can imagine at least 3 key specifications:
>
> Pull-- If the control voltage is varied from minimum to maximum, how much
> does the frequency change?
>

This is very important in our application.  If unable to vary frequency
enough, a lock is unattainable.  This translates to either a "muted"
signal, or full-blown digital noise.

> Accuracy-- If a particular control voltage is applied, how close is the
> frequency to the expected value?
>

It doesn't really matter (to an extent) because the osc. is being used to
lock onto an existing clock.  It is not being used as an absolute
"reference" persay.

> Stability-- With the control voltage held constant, how much does the
> frequency change due to other factors?  Poor stability of course implies
> poor accuracy.
>

Again, as long as the circuit continues to oscillate, it's okay.

> If you need a circuit with large pull and high accuracy, it's going to be
> expensive.  Most likely you don't, especially if a constant input signal
> is present that your circuit can phase or frequency lock to.
>

That is correct.  But because there are many input sources to choose from,
high pull IS necessary, as accuracy and stability are not so important.

<SNIP>

>
> Without knowing about your application, I couldn't say if various
> "hybrid" techniques could be used.  For example, an oscillator with poor
> accuracy could be periodically re-calibrated to a fixed crystal
> oscillator.  Or the pull of an oscillator could be increased by using
> digital techniques to modify the output frequency.  A "pretty good
> feedback circuit" can be very useful.  Feedback can make many
> "impossible" things possible.
>

Could you elaborate on the former hybrid?

-Vince

>
>
>
> ___________________________________________________________________
> Get the Internet just the way you want it.
> Free software, free e-mail, and free Internet access for a month!
> Try Juno Web: dl.http://www.juno.com/dynoget/tagj.
>

1999\08\12@121953 by Vincent Deno

flavicon
face
Could you please write with the specifications of the single-crystal
design?  Thus far, we have found the crystall oscillators perform much
better.  FYI, I am Vince.  Dave Kerstetter is the main Design Engineer.
He is still exploring other avenues and is extremely busy with several
projects.

-Vince Deno


{Quote hidden}

More... (looser matching)
- Last day of these posts
- In 1999 , 2000 only
- Today
- New search...