Searching \ for '[PIC]: use VCC for crystal bypass caps?' 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/microchip/devices.htm?key=pic
Search entire site for: 'use VCC for crystal bypass caps?'.

Exact match. Not showing close matches.
PICList Thread
'[PIC]: use VCC for crystal bypass caps?'
2002\10\03@014901 by Bill Westfield

face picon face
This came up in a beginner thread, but I'd like to get a definitive answer.
Can I use VCC for the "ground" side of the crystal's phase shift caps?
Theory says VCC is equivilent to ground from an AC perspective, and it
would certainly make most PCB layouts a bit cleaner, but I'd like to
hear from someone who has done this regularly and thinks that it actually
works!

Thanks
Bill W

--
http://www.piclist.com hint: To leave the PICList
spam_OUTpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu


2002\10\03@035655 by Jinx

face picon face
> This came up in a beginner thread, but I'd like to get a definitive
> answer. Can I use VCC for the "ground" side of the crystal's
> phase shift caps? Theory says VCC is equivilent to ground from
> an AC perspective, and it would certainly make most PCB layouts
> a bit cleaner, but I'd like to hear from someone who has done this
> regularly and thinks that it actually
> works!
>
> Thanks
> Bill W

The easy way would be to try it and see what happens. I don't recall
ever seeing a micro or (parallel) oscillator that used caps to Vcc,
apart from that one Microchip circuit. Microchip's own guide to PIC
oscillators

http://www.microchip.com/download/appnote/rfpic/00826a.pdf

has no mention that I can see of caps to Vcc, and all diagrams show
caps to ground. So does the Mid-range Manual, and several other MC
pdfs I looked through. Maybe caps to Vcc is acceptable, but putting
them to ground is certainly the prevailing method. Mention is made in
the M-RM of the quiescent charging of caps during SLEEP and how
that can make oscillator start-up difficult. Perhaps that's a clue

--
http://www.piclist.com hint: To leave the PICList
.....piclist-unsubscribe-requestKILLspamspam@spam@mitvma.mit.edu


2002\10\03@043131 by Russell McMahon

face
flavicon
face
> This came up in a beginner thread, but I'd like to get a definitive
answer.
> Can I use VCC for the "ground" side of the crystal's phase shift caps?
> Theory says VCC is equivilent to ground from an AC perspective, and it
> would certainly make most PCB layouts a bit cleaner, but I'd like to
> hear from someone who has done this regularly and thinks that it actually
> works!

It works.
The books generally say not to do it. The reasons given often don't apply.
They SAY that you will introduce noise into the oscillator (which you may).
I don't see that as a bad thing in many cases. You are liable to get a
degree of clock jitter or "spreading" and this is liable to reduce the peak
harmonic intensity. The effect is liable to be small. Use of varicaps as the
crystal caps may be indicated :-).

For applications requiring high clock stability or purity this may not be a
good scheme but for most real world applications it seems fine.

Jinx notes that startup may be affected - this is an area of the blackest of
magics and if you have startup problems this may make them worse - or
better.

I use the arrangement for circuits on breadboard or Vero board (US = vector
board ?).


       Russell McMahon.

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestspamKILLspammitvma.mit.edu


2002\10\03@044445 by Michael Rigby-Jones

flavicon
face
> -----Original Message-----
> From: Jinx [SMTP:.....joecolquittKILLspamspam.....CLEAR.NET.NZ]
> Sent: Thursday, October 03, 2002 8:57 AM
> To:   EraseMEPICLISTspam_OUTspamTakeThisOuTMITVMA.MIT.EDU
> Subject:      Re: [PIC]: use VCC for crystal bypass caps?
>
> > This came up in a beginner thread, but I'd like to get a definitive
> > answer. Can I use VCC for the "ground" side of the crystal's
> > phase shift caps? Theory says VCC is equivilent to ground from
> > an AC perspective, and it would certainly make most PCB layouts
> > a bit cleaner, but I'd like to hear from someone who has done this
> > regularly and thinks that it actually
> > works!
> >
> > Thanks
> > Bill W
>
There is no reason why it shouldn't work, but ISTR this was discussed quite
extensively a long time ago on the list and the conclusion was that it could
potentialy cause bad EMC problems.  The prefered solution is to run the caps
directly back to the PIC's ground pin.

Regards

Mike

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestspamspam_OUTmitvma.mit.edu


2002\10\03@074554 by Olin Lathrop

face picon face
> This came up in a beginner thread, but I'd like to get a definitive
answer.
> Can I use VCC for the "ground" side of the crystal's phase shift caps?
> Theory says VCC is equivilent to ground from an AC perspective, and it
> would certainly make most PCB layouts a bit cleaner, but I'd like to
> hear from someone who has done this regularly and thinks that it actually
> works!

In theory, you are right, Vcc is supposed to be equivalent to ground from an
AC perspective.  However the crystal driver circuit almost certainly uses
ground as a reference when it needs one.  This means that the AC signal will
have to find its way back to the ground of the crystal driver somehow.  It
will have to take a longer higher impedence path if you connect the caps to
Vcc.

Personally I wouldn't do this because it increases the effective length of
the cap connections, and high frequency noise on Vcc (there is always some)
will be coupled into the crystal signal.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com hint: To leave the PICList
@spam@piclist-unsubscribe-requestKILLspamspammitvma.mit.edu


2002\10\03@101114 by Robert E. Griffith

flavicon
face
While we are talking about oscillators...  Do you sometimes put a 1M
resistor in parallel with the crystal?  I am using a STA013 MP3 decoder that
needs a 14.7456 oscillator.  The datasheet just says that it uses an
industry standard oscillator.  I have seen a sample circuit however, that
uses the crystal, two caps to ground, and a 1M resistor in parallel.

So do you why someone might do this?

--BobG

{Original Message removed}

2002\10\03@112858 by Bob Blick

face picon face
Hi Bill,

If the logic uses ground as a reference it would be best to use ground as
the return for the caps. Otherwise you will introduce jitter in the
oscillator, which may or may not be a problem depending on how much noise
there is on V+.

However the PIC is CMOS so the reference, per say, is divided between the
rails. That division is not equal, though. Note that the input threshold
for most pins is "HCT-like" and closer to ground than V+. That says to me
that using ground as the return for the caps is better than V+ for
minimizing clock jitter.

My conclusion is that neither is perfect but that ground is less
imperfect.

Cheerful regards,

Bob

--
http://www.piclist.com hint: To leave the PICList
KILLspampiclist-unsubscribe-requestKILLspamspammitvma.mit.edu


2002\10\03@114228 by Spehro Pefhany

picon face
At 10:08 AM 10/3/02 -0400, you wrote:
>While we are talking about oscillators...  Do you sometimes put a 1M
>resistor in parallel with the crystal?  I am using a STA013 MP3 decoder that
>needs a 14.7456 oscillator.  The datasheet just says that it uses an
>industry standard oscillator.  I have seen a sample circuit however, that
>uses the crystal, two caps to ground, and a 1M resistor in parallel.
>
>So do you why someone might do this?

Yes, the oscillator uses a CMOS inverter that must be biased into the
linear region in order to make an oscillator. Some chips have this
relatively high-value resistor internally, and others require it to
be supplied externally, giving a bit more control over the value,
especially if it should be high (very high values are hard to make
on chips).  PICs have the resistor internally.

On the original topic, I've seen recommendations to replace the capacitors
with two equal valued capacitors (each) of half the total value bridging
the supply. That should represent an improvement in an ideal world,
as the circuit is symmetrical wrt the power supplies. Personally, I think
if there's enough noise to interfere with the clock significantly,
that should be addressed, and if it's a cheap cheap product then
the cost of the extra caps won't be nice either.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
RemoveMEspeffTakeThisOuTspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestspamBeGonespammitvma.mit.edu


2002\10\03@123352 by Wagner Lipnharski

flavicon
face
William Chops Westfield wrote:
> This came up in a beginner thread, but I'd like to get a definitive
> answer. Can I use VCC for the "ground" side of the crystal's phase
> shift caps? Theory says VCC is equivilent to ground from an AC
> perspective, and it would certainly make most PCB layouts a bit
> cleaner, but I'd like to hear from someone who has done this
> regularly and thinks that it actually works!
>
> Thanks
> Bill W
>
> --
> http://www.piclist.com hint: To leave the PICList
> TakeThisOuTpiclist-unsubscribe-requestEraseMEspamspam_OUTmitvma.mit.edu

Hi Bill,
In theory it really doesn't matter where you anchor your AC signals. Ground
or VCC can act the same.

I just see two big issues.

1) Normally we use to lay out larger real estate copper for ground. Not
only this, but normally your Ground will be connected to other Grounds
around, what will create a larger ballast and will spread even more your
Ground potential to suck down noise and abnormal energy.  What this mean?
in a direct path, Ground will always be the easier way for those signals to
travel.  By other side, VCC as usual designed, uses not so heavy copper
rails, a VCC track of 1/8" wide with so many turns and twists can represent
a higher impedance for a high frequency point of view.

2) A blow up capacitor with its imminent short circuit, from output
oscillator clock pin to ground, is not a big issue, but try to do the same
to VCC and you will have a fried oscillator faster than you can say
"oops!".

Most microchips try to balance their internal low impedance much more to
ground than to VCC.  Even discrete components designs, you can see much
more component leads to ground than to VCC. This, shows a natural tendency
to a lower impedance in general to ground than VCC.

Of course the large and small capacitors at any power supply offer a low
impedance to frequencies, so, for a decoupling capacitor, it really doesn't
matter much where you anchor it, if, or course, the path is free and direct
impedance is low for such frequency.

Perhaps one problem to consider is that the return path via VCC will be
always longer than via Ground, this can create unexpected issues and
intermittent problems.

/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
Wagner Lipnharski - UST Research Inc
Orlando FLorida - USA - http://www.ustr.net
/_/_/_/ Atmel AVR Consultant /_/_/_/

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspamTakeThisOuTmitvma.mit.edu


2002\10\03@134711 by Olin Lathrop

face picon face
> While we are talking about oscillators...  Do you sometimes put a 1M
> resistor in parallel with the crystal?  I am using a STA013 MP3 decoder
that
> needs a 14.7456 oscillator.  The datasheet just says that it uses an
> industry standard oscillator.  I have seen a sample circuit however, that
> uses the crystal, two caps to ground, and a 1M resistor in parallel.
>
> So do you why someone might do this?

Some oscillator circuits that are essentially inverters (180 degree phase
shift) need this for DC feedback.  It keeps the DC levels near the center of
the range.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestEraseMEspam.....mitvma.mit.edu


2002\10\03@145219 by Robert E. Griffith

flavicon
face
Thanks Olin and Spehro for clearing that up.

This Zilog app note on oscillators,
http://www.zilog.com/docs/z8/appnotes/on_chip.pdf, suggests a specific
layout that encloses the crystal and caps in a ground trace (sort of).  It
would be tough for me to use the space that they suggest, so I have just
stacked the crystal, 1M resistor and caps close to the oscillator pins and
circled them with a ground trace that picks up both caps and returns to the
chip's ground pin which is right next to the oscillator pins.  I was not
sure whether to connect the ground trace to the chip's ground pin, (which is
connected to the ground plane through a via on the other side), or put a via
to ground at the top of the components which would make it symmetrical wrt
the two caps.  I also have I feeling that I might be overdoing it - that it
does not make much difference either way.  What do you think?

--BobG



{Original Message removed}

2002\10\03@150410 by Dale Botkin

flavicon
face
On Thu, 3 Oct 2002, Robert E. Griffith wrote:

> I also have I feeling that I might be overdoing it - that it
> does not make much difference either way.  What do you think?

Go with your feeling.

Dale

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspammitvma.mit.edu


2002\10\03@152415 by Spehro Pefhany

picon face
At 02:49 PM 10/3/02 -0400, you wrote:
>Thanks Olin and Spehro for clearing that up.
>
>This Zilog app note on oscillators,
>http://www.zilog.com/docs/z8/appnotes/on_chip.pdf, suggests a specific
>layout that encloses the crystal and caps in a ground trace (sort of).  It
>would be tough for me to use the space that they suggest, so I have just
>stacked the crystal, 1M resistor and caps close to the oscillator pins and
>circled them with a ground trace that picks up both caps and returns to the
>chip's ground pin which is right next to the oscillator pins.  I was not
>sure whether to connect the ground trace to the chip's ground pin, (which is
>connected to the ground plane through a via on the other side), or put a via
>to ground at the top of the components which would make it symmetrical wrt
>the two caps.  I also have I feeling that I might be overdoing it - that it
>does not make much difference either way.  What do you think?

Avoid running other traces through the clock oscillator parts. The way the
ground gets back is a bit more dependent on the exact circuit and what all
is on the board. I'd "look" at how currents flow through the ground plane
(4-layer board?) before answering that one. You don't want current
flowing from elsewhere on the board going to clock capacitor ground, so
a wide direct trace that goes a very short distance and NOWHERE ELSE is
best. Also, keep all the traces close to each other to minimize EMC
issues. I routinely do boards with amperes and microvolts on the same
PCB, it's not that hard if you can visualize the current flows.

Overdoing it? Maybe, but once you get an idea of what is right, it will
become second nature.

P.S. Don't forget that through hole crystal cases are typically metal and
conduct. I like to see the crystal case grounded if possible (which is a
compatible layout with a ceramic resonator if the crystal is made with
a ground pin). Also you should have a routing keepout on the top layer
around the crystal case outline rather than depending on the solder
mask as insulation.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
RemoveMEspeffEraseMEspamEraseMEinterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspam_OUTspamKILLspammitvma.mit.edu


2002\10\03@154635 by Rick C.

flavicon
face
William Chops Westfield wrote:
> This came up in a beginner thread, but I'd like to get a definitive
> answer. Can I use VCC for the "ground" side of the crystal's phase
> shift caps? Theory says VCC is equivilent to ground from an AC
> perspective, and it would certainly make most PCB layouts a bit
> cleaner, but I'd like to hear from someone who has done this
> regularly and thinks that it actually works!
>
I do this often to facilitate a clean layout. Just make sure you have a good
0.1uf cap nearby that Vcc point to ground and as close to the PIC as possible
to reduce any *possible* noise.

At 10:08 AM 10/3/02 -0400, you wrote:
>While we are talking about oscillators...  Do you sometimes put a 1M
>resistor in parallel with the crystal?  I am using a STA013 MP3 decoder that
>needs a 14.7456 oscillator.  The datasheet just says that it uses an
>industry standard oscillator.  I have seen a sample circuit however, that
>uses the crystal, two caps to ground, and a 1M resistor in parallel.
>
>So do you why someone might do this?

The Mid-Range MCU Reference Manual discusses this in detail in Section 2.3.
Briefly, there is an internal feedback resistor in most PICs which *should* be
adequate in most applications. A scope on the oscillator output pin will tell
all. The waveform should be somewhat symmetrical and swing evenly from Vcc to
Gnd and not hitting either. Fig 2.2 in the manual shows this. If you happen to
have a crystal with too high of a Q, the waveform will show clipping or
saturation of the output. This will cause an unstable oscillation and could
possibly physically fracture the crystal. If this is the case, an external
feedback resistor (between 1m and 10m) across the crystal will tame it.
Rick

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestTakeThisOuTspamspammitvma.mit.edu


2002\10\03@200854 by Bill Westfield

face picon face
   1) Normally we use to lay out larger real estate copper for
   ground. Not only this, but normally your Ground will be connected
   to other Grounds around, what will create a larger ballast and
   will spread even more your Ground potential to suck down noise
   and abnormal energy.  What this mean?  in a direct path, Ground
   will always be the easier way for those signals to travel.  By
   other side, VCC as usual designed, uses not so heavy copper
   rails, a VCC track of 1/8" wide with so many turns and twists can
   represent a higher impedance for a high frequency point of view.

Surely your bypass caps short VCC to your heavy ground plane "very close"
to the PIC crystal anyway, at least for the sake of AC (and the crystal
AC will be the highest freq on the board, right?)

BillW

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspamspamspamBeGonemitvma.mit.edu


2002\10\04@024301 by ISO-8859-1?Q?Ruben_J=F6nsson?=

flavicon
face
> Surely your bypass caps short VCC to your heavy ground plane "very
> close" to the PIC crystal anyway, at least for the sake of AC (and the
> crystal AC will be the highest freq on the board, right?)
> The crystal AC will not be the highest freq on the board. This is a, more or less, sinusoidal signal but it gets converted to square wave signals inside the processor which generates a lot of overtones.

Ruben
==============================
Ruben Jönsson
AB Liros Electronic
Box 9124, 200 39 Malmö, Sweden
TEL INT +46 40142078
FAX INT +46 40947388
RemoveMErubenKILLspamspampp.sbbs.se
==============================

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\04@032201 by Morgan Olsson

picon face
One important thing is to route the xtal caps to the PIC GND (or Vcc) pin, not "any" nearby GND/Vcc track.

Thus minimizing the current loop (the xal driver is in the PIC right, so dump the currents as close to the driver as possible.  Btw also you Vcc-GND bypass cap is close to PIC, right, so both + and - side of driver is close connected to xtal caps.

Thus minimizong 20MHz (whatever) frequency induced in GND or Vcc trace.

/Morgan

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\04@033407 by Morgan Olsson

picon face
>Surely your bypass caps short VCC to your heavy ground plane "very close"
>to the PIC crystal anyway, at least for the sake of AC

Still, as the OSC pins is placed close to Vcc/GND it is easy to route the xtal caps to PIC pins separately.  (this is also the reason they are placed there!)  This will minimize radiation, and also make the osc less suspect to pick up any peaks on supply rails (which could make the PIC execute the cirrent operation erroneously)

>(and the crystal AC will be the highest freq on the board, right?)

Wrong.  Possibly the highest "fundamental frequency" but it is a nice sine, and also low power.  The hard switching in the many CPU transistors make lot stronger noise of much higher freqencies.  Also a hard switching output or i.e a not so well designed motor PWM driver may ring at high freqency every on/off change.

/Morgan

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...