Searching \ for '[PIC] Max speed from 887' 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: 'Max speed from 887'.

Exact match. Not showing close matches.
PICList Thread
'[PIC] Max speed from 887'
2008\06\17@125634 by Tomás Ó hÉilidhe

picon face

I'm used to using the 16F684, and I had the following macro for getting
the greatest speed out of it:

   #define CrankTheSpeed() ((void)(OSCCON |= 0b1110000,OSCTUNE = 0b1111))

I got myself a "Pickit2 Debug Express" today though, and from now on
I'll be working with the 16F887.

I took a quick glance at the datasheet for the 887 and it tells me that
the max speed I'll get is 8 MHz if I choose to use internal oscillators.

I want the absolute maximum speed though (so I think that means I use an
RC circuit to tell it what speed I want).

Could someone please explain to me what pin I put the RC on, what RC
values to use, what registers to set, etc.?

2008\06\17@131526 by James Nick Sears

flavicon
face
To get full 20MHz, I would just get a powered oscillator and use EC
mode.  Something like
search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=X119-ND
(That one is for 5V supply, you can get 3.3V capable ones as well if
you need)

If you're not concerned with a few milliamps of extra power
consumption and an extra dollar or two of cost, it will probably save
you some headaches.

-n.



On Tue, Jun 17, 2008 at 12:56 PM, Tomás Ó hÉilidhe <spam_OUTtoeTakeThisOuTspamlavabit.com> wrote:
{Quote hidden}

>

2008\06\17@132354 by olin piclist

face picon face
Tomás Ó hÉilidhe wrote:
> Could someone please explain to me what pin I put the RC on, what RC
> values to use, what registers to set, etc.?

Have you even looked at the datasheet?  There is a section in there that
specifically covers this.

By the way, I'd use a crystal or ceramic resonator at 20MHz, not a external
R/C.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\17@133549 by olin piclist

face picon face
James Nick Sears wrote:
> To get full 20MHz, I would just get a powered oscillator and use EC
> mode.

How is that substantially easier than a crystal and two caps or a single
ceramic resonator with internal caps?  Those canned oscillators are bulky,
expensive, use a lot of power, and glitch the supply.  At the minimum you
need two parts, the oscillator and the bypass cap.

About the only advantages are that is uses one less PIC pin and some
oscillators can be more accurate than just a crystal.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\17@134125 by Michael Rigby-Jones

picon face


> -----Original Message-----
> From: .....piclist-bouncesKILLspamspam@spam@mit.edu [piclist-bouncesspamKILLspammit.edu] On
Behalf
{Quote hidden}

Because you don't need to jump through hoops to get the correct value of
the caps for the crystal you have which you can't find a datasheet for?
Neither do you have to worry about start-up issues.

Canned oscillators are certainly very power hungry, but they have their
uses.

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2008\06\17@134533 by Michael Rigby-Jones

picon face


{Quote hidden}

Using an external RC oscillator for high speeds is going to give you an unpredictable frequency that will be very prone to drifting with temperature and hand capacitance effects.  Don't do it.

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2008\06\17@140003 by M. Adam Davis

face picon face
On 6/17/08, Olin Lathrop <olin_piclistspamspam_OUTembedinc.com> wrote:
> Tomás Ó hÉilidhe wrote:
> > Could someone please explain to me what pin I put the RC on, what RC
> > values to use, what registers to set, etc.?
>
> Have you even looked at the datasheet?  There is a section in there that
> specifically covers this.

Specifically section 3.4.4 shows the schematic, and describes the
factors that determine the frequency.  Unfortunately they do not
provide an equation, on some data sheets you can find a table or graph
which gives an idea of where to start if you're looking for
aparticular frequency, but I did not find one in this datasheet (in
the time I am able to spend right now).

You might check out the 16f84 data sheet:
http://ww1.microchip.com/downloads/en/devicedoc/35007b.pdf

The figures 10-7, 10-8, and 10-9 give graphs of RC oscillator
frequency versus supply voltage for a given C and R at room
temperature.  While the internal oscillator design may be slightly
different in the chip you're using, it should be a good starting
point, and is likely to have similar characteristics.

Good luck!

-Adam

--
EARTH DAY 2008
Tuesday April 22
Save Money * Save Oil * Save Lives * Save the Planet
http://www.driveslowly.org

2008\06\17@140526 by James Nick Sears

flavicon
face
The advantage is that clearly the OP has enough variables to deal
with.  I've had enough problems with oscillators starting up over my
years of projects that I would recommend in that case eliminating a
few of those variables.  If the OP was in a position to want to learn
about the intricacies of oscillator design I doubt he would have asked
his question in the first place.

Olin, I don't know how much you value your time, but I wouldn't be
worrying about a $1.85 part on a one-off if it saves me more than
about a minute.

The best part about this is that I KNEW this post was coming from Olin
before I even sent my response.  Beginners shouldn't ask for help, but
also shouldn't use parts that will save them trouble and minimize the
need to ask for help.  Olin, I think maybe you should buy a truckload
of transistors and work on building your own PIC from scratch.  Just
to prove how hardcore you really are.



On Tue, Jun 17, 2008 at 1:30 PM, Olin Lathrop <@spam@olin_piclistKILLspamspamembedinc.com> wrote:
{Quote hidden}

> -

2008\06\17@140850 by James Nick Sears

flavicon
face
And I knew this one was coming as well.  I should just write an Olin
simulator.  Couldn't be more than a few hundred lines of code.

And Olin, have YOU even looked at the datasheet?  There is some info
yes, but it's not immediately apparent how to select RC values or the
range of appropriate frequencies with an RC to someone who doesn't
have some prior knowledge of how RC oscillators work.

And yes, you can say the OP is a fool and should learn about RC
oscillators and analog circuits and --

But to that, I would say:

Go buy an oscillator and build whatever it is you want to build.



On Tue, Jun 17, 2008 at 1:26 PM, Olin Lathrop <KILLspamolin_piclistKILLspamspamembedinc.com> wrote:
{Quote hidden}

>

2008\06\17@143708 by olin piclist

face picon face
James Nick Sears wrote:
> The best part about this is that I KNEW this post was coming from Olin
> before I even sent my response.  Beginners shouldn't ask for help, but
> also shouldn't use parts that will save them trouble and minimize the
> need to ask for help.  Olin, I think maybe you should buy a truckload
> of transistors and work on building your own PIC from scratch.  Just
> to prove how hardcore you really are.

I guess you missed the point that a ceramic resonator is *easier* than a
canned oscillator.  You hook it up and you're done.  I've seen canned
oscillators cause trouble in circuits, usually because someone didn't pay
attention to bypassing and high frequency loop currents.  In other words,
canned oscillators are more advanced and require more expertise to get
right.  They look easy, but they're really not.  They are not for beginners.

On the other hand, I don't remember having a problem with a PIC crystal
oscillator.  My experience is that they just work.  If you want to use a
junkbox crystal that you can't identify to find the datasheet, just use 22pF
caps.  The frequency might be a little off, but it will oscillate.

I don't use ceramic resonators much, but the ones with the built in load
caps are the easiest of all.  Connect the end leads to the PIC pins and the
center lead to ground and you're all done.

No matter what you chose, you still have to set up the PIC oscillator mode,
that's a wash.  For 20MHz on a 16F887 there is no doubt this needs to be HS
for a crystal or resonator.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\17@144408 by olin piclist

face picon face
James Nick Sears wrote:
> There is some info
> yes, but it's not immediately apparent how to select RC values or the
> range of appropriate frequencies with an RC to someone who doesn't
> have some prior knowledge of how RC oscillators work.

Take a look at what he really asked:

>> Could someone please explain to me what pin I put the RC on, what RC
>> values to use, what registers to set, etc.?

Maybe the RC values might not be obvious, but the pin to use!!?  This guy
didn't bother to look.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\17@145247 by Stephen D. Barnes

flavicon
face
James Nick Sears wrote:
> And I knew this one was coming as well.  I should just write an Olin
> simulator.  Couldn't be more than a few hundred lines of code.
>
> And Olin, have YOU even looked at the datasheet?
> <clip>

This may be OT but, I just have to speak up! I, and I hope others as
well, appreciate the consistent, and accurate info that Olin provides.
He has attempted to provide useful info concerning th OP's question
while minimizing issues brought about by different oscillator
strategies. He also has not engaged in any name calling. Please don't
attempt to lure anyone into another flame war.
Thanks

--
Regards,
Stephen D. Barnes

2008\06\17@150008 by Jan-Erik Soderholm

face picon face
Tomás Ó hÉilidhe wrote:

> I'm used to using the 16F684, and I had the following macro for getting
> the greatest speed out of it:
>
>     #define CrankTheSpeed() ((void)(OSCCON |= 0b1110000,OSCTUNE = 0b1111))
>
> I got myself a "Pickit2 Debug Express" today though, and from now on
> I'll be working with the 16F887.
>
> I took a quick glance at the datasheet for the 887 and it tells me that
> the max speed I'll get is 8 MHz if I choose to use internal oscillators.
>
> I want the absolute maximum speed though (so I think that means I use an
> RC circuit to tell it what speed I want).
>
> Could someone please explain to me what pin I put the RC on, what RC
> values to use, what registers to set, etc.?
>

Exactly *why* do you want a RC osc (or the INTOSC)
instead of the usual crystal+caps ??

And, btw, the 684 and the 887 are similar
when it comes to OSCCON and OSCTUNE, so no
difference there. You'll get more or less
the same performance from both.

If you need more performance with *no* external
RC or crystals, use one of the PIC18's with
8 Mhz INTOSC and 4*PLL for a 32 Mhz effective
*internal* oscillator.

> > Could someone please explain to me what pin I put the RC on...

*That*, and the other related qustions, are clearly an
area where the datasheet is the *PERFECT* source for
the answer.


Regards,
Jan-Erik.

2008\06\17@150315 by Spehro Pefhany

picon face
Quoting Olin Lathrop <RemoveMEolin_piclistTakeThisOuTspamembedinc.com>:

> James Nick Sears wrote:
>> There is some info
>> yes, but it's not immediately apparent how to select RC values or the
>> range of appropriate frequencies with an RC to someone who doesn't
>> have some prior knowledge of how RC oscillators work.
>
> Take a look at what he really asked:
>
>>> Could someone please explain to me what pin I put the RC on, what RC
>>> values to use, what registers to set, etc.?
>
> Maybe the RC values might not be obvious, but the pin to use!!?  This guy
> didn't bother to look.

In addition, you cannot get maximum fosc from a PIC using the RC oscillator
and stay within the recommended range of values, so it's a not a good  
way to go if you really want "max speed" as in the subject line.


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


2008\06\17@150544 by James Nick Sears

flavicon
face
On Tue, Jun 17, 2008 at 2:39 PM, Olin Lathrop <TakeThisOuTolin_piclistEraseMEspamspam_OUTembedinc.com> wrote:
> I guess you missed the point that a ceramic resonator is *easier* than a
> canned oscillator.  You hook it up and you're done.  I've seen canned
> oscillators cause trouble in circuits, usually because someone didn't pay
> attention to bypassing and high frequency loop currents.  In other words,
> canned oscillators are more advanced and require more expertise to get
> right.  They look easy, but they're really not.  They are not for beginners.

Fair enough, but my experience differs.  I spent the majority of the
time on my first PIC project for hire some years back tinkering with a
crystal oscillator circuit that wouldn't start reliably.  This was
with an SMT PCB layout and traces about .2" long and much
experimentation with cap values.  I've since had projects that worked
fine with crystals of course, but I've also never had a problem with a
canned osc and neither has any of the dozens of people I've seen use
them in projects.  Lucky I guess.

{Quote hidden}

> -

2008\06\17@151330 by Jan-Erik Soderholm

face picon face
James Nick Sears wrote:

> The advantage is that clearly the OP has enough variables to deal
> with.  I've had enough problems with oscillators starting up over my
> years of projects that I would recommend in that case eliminating a
> few of those variables.  If the OP was in a position to want to learn
> about the intricacies of oscillator design I doubt he would have asked
> his question in the first place.

He asked specificaly about a switch from the 684 to the 887,
which actualy has no changes in regard to the oscillator configs.

{Quote hidden}

You would *not* have written that if if didn't happend to
be a comment on a post from Olin. Please cool down a little...

Best Regards,
Jan-Erik.

2008\06\17@151427 by James Nick Sears

flavicon
face
Look, I'm not attempting to lure anyone into anything.  If I'm the
only one who thinks the attitude toward beginners here is a little
much and needs someone pulling in the other direction, then I'm
clearly in the wrong place and I apologize and will move on to a
different community that actually values people who are trying to get
things done with technology, and not only people who are obsessed with
the technology itself.  I'm not asking anyone to moderate, or anyone
to get banned or any of that crap, I'm just trying to provide a little
balance.  I think Olin can handle it.  If he can't, then he REALLY
should revisit the way he treats other people.  But I'm confident he
can and that everything will be just fine.



On Tue, Jun 17, 2008 at 2:51 PM, Stephen D. Barnes
<RemoveMEstephendbarnesspamTakeThisOuTcavtel.net> wrote:
{Quote hidden}

> -

2008\06\17@151449 by Herbert Graf

flavicon
face
On Tue, 2008-06-17 at 14:39 -0400, Olin Lathrop wrote:
> I guess you missed the point that a ceramic resonator is *easier* than a
> canned oscillator.  You hook it up and you're done.  I've seen canned
> oscillators cause trouble in circuits, usually because someone didn't pay
> attention to bypassing and high frequency loop currents.  In other words,
> canned oscillators are more advanced and require more expertise to get
> right.  They look easy, but they're really not.  They are not for beginners.
>
> On the other hand, I don't remember having a problem with a PIC crystal
> oscillator.  My experience is that they just work.  If you want to use a
> junkbox crystal that you can't identify to find the datasheet, just use 22pF
> caps.  The frequency might be a little off, but it will oscillate.

My experience is the exact opposite. While canned oscillators have been
100% for me (even surviving insertion backwards), crystals are another
story.

While in general the PIC XTAL driver seems to be above average, I have
had issues where everything is done supposedly correctly, yet the
crystal either doesn't start up, or runs erratically.

To be fair all problems were on prototype circuits, usually soldered
breadboard. Of course, this is EXACTLY the type of circuit construction
someone just starting out with PICs will likely use, so my experience is
relevant.

FWIW I've never had a problem with lower frequency crystals (say in the
4-8MHz range), but 16-20MHz HAVE been a problem, especially with the
older 16F parts (for example, upgrading a design from an 877 to an 877a
part resulted in no oscillation). I haven't experienced a problem with
the newer parts, but then the sample space is much smaller since either
I use the internal RC, or use a lower crystal with the 4x PLL enabled.

Outside of the PIC world I pretty much always either use an internal
oscillator if available, or I use a lower crystal and PLL my way up to
the speeds I want. Just reduces the headache.

TTYL

2008\06\17@154221 by Bob Blick

face
flavicon
face

On Tue, 17 Jun 2008 17:56:19 +0100, "Tomás Ó hÉilidhe" <toeEraseMEspam.....lavabit.com>
said:

>
> I took a quick glance at the datasheet for the 887 and it tells me that
> the max speed I'll get is 8 MHz if I choose to use internal oscillators.

Yes.

> I want the absolute maximum speed though (so I think that means I use an
> RC circuit to tell it what speed I want).

Probably not. RC mode is not great at very high frequencies and uses a
lot of current. But it varies a lot from part number to part number, so
you should probably dig deeply in the datasheet if you really want to
use RC mode. And find that info is incomplete or preliminary. Nobody
uses RC mode on the '887.

> Could someone please explain to me what pin I put the RC on, what RC
> values to use, what registers to set, etc.?

I doubt you'll find anyone here who has used RC mode on an '887. Use a
crystal or approved ceramic resonator(I posted resonator tests a while
back, some don't work so well at 20MHz).

Cheerful regards,

Bob

--
http://www.fastmail.fm - I mean, what is it about a decent email service?

2008\06\17@154439 by James Nick Sears

flavicon
face
> You would *not* have written that if if didn't happend to
> be a comment on a post from Olin. Please cool down a little...

Consider me cool.  Clearly the consensus here is against my take on
the situation, so no more defending other people.  If you haven't read
the datasheet cover to cover, be damned.  But also consider it noted
that I'm not personally putting up with anyone's s***, particularly
from people who have shown a history of consistently dishing it out.

I stayed out of the whole Olin moderation debate, not through any
personal willpower, but because frankly I don't care whatsoever.  But
I asked my first PIC question in a few years right after that and
immediately saw the cause for all the uproar: "Why are you
prototyping, stupid?!?!?!" etc.

Again, I'm not trying to start anything, continue anything, or do
anything but end this, but don't be surprised when someone disrespects
me and I react.  That's a pattern that you can expect to continue.

That's my final word on the topic.  Back to work everyone.

-n.

2008\06\17@155917 by Wouter van Ooijen

face picon face
> Look, I'm not attempting to lure anyone into anything.  If I'm the
> only one who thinks the attitude toward beginners here is a little
> much and needs someone pulling in the other direction, then I'm
> clearly in the wrong place and I apologize and will move on to a
> different community that actually values people who are trying to get
> things done with technology, and not only people who are obsessed with
> the technology itself.  

I value people who are to the point and accurate, and try to direct
people in the direction where answers can be found (often worded better
than anyone could in a quick email) instead of giving the answer itself.

And I hate it when people who act in that spirit are bashed.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\06\17@160202 by Jan-Erik Soderholm

face picon face
James Nick Sears wrote:

> If you haven't read
> the datasheet cover to cover, be damned.

I do not know about "cover to cover", the
actual part about the RC oscillator would
probably be enough. I do not think it's even
a full page. Actualy (just checked) it
less then half a page on page 65. It clearly
specifies OSC1/CLKIN as the pin to use.


> Back to work everyone.

Absolutly !

:-) :-)

Jan-Erik.

2008\06\17@161849 by James Nick Sears

flavicon
face
On Tue, Jun 17, 2008 at 3:58 PM, Wouter van Ooijen <EraseMEwouterspamvoti.nl> wrote:
> I value people who are to the point and accurate, and try to direct
> people in the direction where answers can be found (often worded better
> than anyone could in a quick email) instead of giving the answer itself.
>
> And I hate it when people who act in that spirit are bashed.


Ooookay.  Not sure what list you've been reading, but that's great.

-n.

2008\06\17@173353 by Tomás Ó hÉilidhe

picon face


James Nick Sears wrote:
> And Olin, have YOU even looked at the datasheet?  There is some info
> yes, but it's not immediately apparent how to select RC values or the
> range of appropriate frequencies with an RC to someone who doesn't
> have some prior knowledge of how RC oscillators work.

I know about RC circuits, about the time constant, about exponential
decay and all that. What I don't know is how the RC values actually
yield a specific processor speed. I just need an equation, something like:

   Processor Speed = 7.56 * Time Constant

And then my next question would be what capacitor value I should use,
given that I can have a really small one with a really big resistance,
or a really big one with a really small resistance.

2008\06\17@173808 by Tomás Ó hÉilidhe

picon face


Bob Blick wrote:
> I doubt you'll find anyone here who has used RC mode on an '887. Use a
> crystal or approved ceramic resonator(I posted resonator tests a while
> back, some don't work so well at 20MHz).

Could you give me a part number for one of these resonator thingies?
I've used oscillators in college, there's basically just a function
generator. I haven't use a resonator before though. I want something
small and cheap.

2008\06\17@174742 by James Nick Sears

flavicon
face
On Tue, Jun 17, 2008 at 5:33 PM, Tomás Ó hÉilidhe <RemoveMEtoeEraseMEspamEraseMElavabit.com> wrote:
>
>
> James Nick Sears wrote:
>> And Olin, have YOU even looked at the datasheet?  There is some info
>> yes, but it's not immediately apparent how to select RC values or the
>> range of appropriate frequencies with an RC to someone who doesn't
>> have some prior knowledge of how RC oscillators work.
>
> I know about RC circuits, about the time constant, about exponential
> decay and all that. What I don't know is how the RC values actually
> yield a specific processor speed. I just need an equation, something like:
>
>    Processor Speed = 7.56 * Time Constant
>
> And then my next question would be what capacitor value I should use,
> given that I can have a really small one with a really big resistance,
> or a really big one with a really small resistance.

OK, to quote Olin, this is in the datasheet, specifically look in the
text at the bottom of Figure 4-5 "External RC Modes"

I believe (this isn't clearly stated in the datasheet AFAIK) that Fosc
= 1/RC, where Fosc(max) for the chip is 20MHz.  The actual instruction
clock is 1/4 this speed, meaning at most you can run 5 million
instructions / second.

However in that figure, MChip recommends C > 20pF and R > 3k, which
yields a max frequency of 16.666MHz.  And as others have mentioned,
pushing the limits of this may not be the most stable arrangement,
hence the recommendations of external oscs, crystals, and ceramic
resonators.

But hey, if you've got the parts, give it a shot and see if it works.

-n.



>
>

2008\06\17@175427 by Bob Blick

face
flavicon
face

On Tue, 17 Jun 2008 22:33:19 +0100, "Tomás Ó hÉilidhe" <RemoveMEtoespam_OUTspamKILLspamlavabit.com>
said:

{Quote hidden}

Below 1 MHz you might be able to do something that simple. But higher
than that the internal(non-linear) resistance of the PIC's RC oscillator
plays too big a part. And stray capacitance becomes a problem when you
get down below 100pF.

I think Microchip used to put charts in the datasheet but not for parts
that have internal oscillators. There's no point.

If you are interested look at a datasheet for an old part like the 16F84
you might find more RC oscillator data. But it is a different chip so
the exact numbers will be different on a 16F887.

Cheerful regards,

Bob

--
http://www.fastmail.fm - A no graphics, no pop-ups email service

2008\06\17@175831 by Wouter van Ooijen

face picon face
> And then my next question would be what capacitor value I should use,
> given that I can have a really small one with a really big resistance,
> or a really big one with a really small resistance.

The proper answer is (if you want reliable maximum speed): none, use a
resonator, crystal, or canned oscillator instead. And read the posts
about those. IMHO a 3-leg resonator (with build-in caps) is the most
hassle-free, but also the least accurate (but for most purposes,
including serail asynch comm, accurate enough).

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\06\17@180442 by Jan-Erik Soderholm

face picon face
Tomás Ó hÉilidhe wrote:
>
> James Nick Sears wrote:
>> And Olin, have YOU even looked at the datasheet?  There is some info
>> yes, but it's not immediately apparent how to select RC values or the
>> range of appropriate frequencies with an RC to someone who doesn't
>> have some prior knowledge of how RC oscillators work.
>
> I know about RC circuits, about the time constant, about exponential
> decay and all that. What I don't know is how the RC values actually
> yield a specific processor speed. I just need an equation, something like:
>
>     Processor Speed = 7.56 * Time Constant
>
> And then my next question would be what capacitor value I should use,
> given that I can have a really small one with a really big resistance,
> or a really big one with a really small resistance.
>

Right. And you have got your answers. The most
common seems to be "Do not do that, unless you
have a realy, realy good reason".

And, from what I know, there isn't any real
answer to your question, RC-osc doesn't give
a particualy predictable oscillator speed.

> Could you give me a part number for one of these
> resonator thingies?

Check :
http://www.voti.nl/shop/catalog.html?xtals
for a couple of examples.

Also note the application notes listed on the
oscillator page (page 64) in the data sheet.

> I want something small and cheap.

What is "small" and what is "cheap" ?
And what about "works" ? :-)

Best Regards,
Jan-Erik.

2008\06\17@180608 by Bob Blick

face
flavicon
face
I'm on the road right now so the best I can do is give you the part
number of the one I use, and it is surface mount:

Manufacturer: ECS
Part Number: ECS-CR2-20.00-B

If you search the piclist archive you may find the list of ones I tested
for compatibility with PICs. But they are all surface mount.

Cheers,

Bob



On Tue, 17 Jun 2008 22:37:55 +0100, "Tomás Ó hÉilidhe" <RemoveMEtoeTakeThisOuTspamspamlavabit.com>
said:
> Could you give me a part number for one of these resonator thingies?
> I've used oscillators in college, there's basically just a function
> generator. I haven't use a resonator before though. I want something
> small and cheap.

--
http://www.fastmail.fm - And now for something completely different…

2008\06\17@181102 by Tomás Ó hÉilidhe

picon face


Wouter van Ooijen wrote:
> The proper answer is (if you want reliable maximum speed): none, use a
> resonator, crystal, or canned oscillator instead. And read the posts
> about those. IMHO a 3-leg resonator (with build-in caps) is the most
> hassle-free, but also the least accurate (but for most purposes,
> including serail asynch comm, accurate enough).

I don't need precision. It can be 18 MHz, it can be 21 MHz. A ball park
figure will do.

Can you please suggest a resonator part number that I can look up?

2008\06\17@181414 by olin piclist

face picon face
Tomás Ó hÉilidhe wrote:
> I know about RC circuits, about the time constant, about exponential
> decay and all that. What I don't know is how the RC values actually
> yield a specific processor speed. I just need an equation, something
> like:
>
>     Processor Speed = 7.56 * Time Constant

If anything it would have to be more the inverse of that.  Think about it.
Longer time constants are going to slow things down, not speed them up.

In any case, there will be a lot of variation and the RC oscillator is not
meant for high speed operation, if it will work at all at the maximum
processor frequency.

> And then my next question would be what capacitor value I should use,
> given that I can have a really small one with a really big resistance,
> or a really big one with a really small resistance.

No, to make higher speeds you will need both smaller resitor and capacitor.

Still, you don't want to use RC for your purpose.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\17@185103 by Rolf

face picon face
Tomás Ó hÉilidhe wrote:
> Wouter van Ooijen wrote:
>  
>> The proper answer is (if you want reliable maximum speed): none, use a
>> resonator, crystal, or canned oscillator instead. And read the posts
>> about those. IMHO a 3-leg resonator (with build-in caps) is the most
>> hassle-free, but also the least accurate (but for most purposes,
>> including serail asynch comm, accurate enough).
>>    
>
> I don't need precision. It can be 18 MHz, it can be 21 MHz. A ball park
> figure will do.
>
> Can you please suggest a resonator part number that I can look up?
>
>  
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=X909-ND

Or, more generally, try search for "20MHz Resonator" at Digikey.com

Rolf

2008\06\17@191014 by Jinx

face picon face
>You might check out the 16f84 data sheet:
> http://ww1.microchip.com/downloads/en/devicedoc/35007b.pdf
>
> The figures 10-7, 10-8, and 10-9 give graphs of RC oscillator

I put some numbers to that

http://home.clear.net.nz/pages/joecolquitt/0f84rc.html

High-speed RC is easily reliable, too many influences that need time
to correct. Time better spent going to the crystal shop

2008\06\17@191426 by Jinx

face picon face
I wrote

> High-speed RC is easily reliable

Should read - High-speed RC is not easily made reliable

2008\06\17@193517 by Apptech

face
flavicon
face
> I don't need precision. It can be 18 MHz, it can be 21
> MHz. A ball park
> figure will do.

> Can you please suggest a resonator part number that I can
> look up?

Gargoyle (OK, just this once, Google) is your friend.

And Digikey is often an immensely useful resource to start
with

       http://www.digikey.com

They are often easier to access data sheets via than many of
the datasheet sites. And they even sell stuff :-).

Typing

       resonator 20 mhz

in their search box returned

       http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=852887;keywords=resonator%2020%20mhz

with almost 40 listings for 20 Mhz resonators.

Also most useful is http://www.findchips .com
but Digikey often sufficeth.

If you use Gargoyle and Digikey first and then ask questions
here with a bit of known background you'll find that the
anciem curmudgeons here become much more helpful and even
downright useful. People like Olin are world famous in
Littleton Massachusetts for being grumpy and abrasive with
people who don't seem to them to have done basic homework.
(It's the seeming that matters :-) ). *However*, if people
are seen to be trying then Olin and co are about the most
stunning PIC resource you can have, bar none. Well worth
cultivating. And there's mainly lurking Scott who seems to
eat and breathe advanced algorithms and tricky ways of doing
things better and faster and cheaper (choose any 3). If you
can ever find a hard enough PIC problem to pose you may even
tempt him out of lurk mode to answer.

Re resonators - if, after looking at eg the Digikey listings
you decide that there are too many choices, then saying that
you have looked at xxx but it's not obvious if there is a
best choice, or what do people prefer etc, then people see
you've had a good attempt at using the available resources
and will usually fall over themselves to help.

Re PIC RC oscillator frequency - AFAIR (not having looked
any time recently) the data sheets have a graph of frequency
versus R & C combos.

You'll find this list a most valuable resource and an
international community, well worth being part of.



regards


       Russell

2008\06\17@194737 by David Meiklejohn

face
flavicon
face
James Nick Sears wrote:
> If I'm the
> only one who thinks the attitude toward beginners here is a little
> much and needs someone pulling in the other direction, then I'm
> clearly in the wrong place and I apologize and will move on to a
> different community that actually values people who are trying to get
> things done with technology, and not only people who are obsessed with
> the technology itself.

I haven't seen any "bad" attitude toward beginners in this thread at all.

If someone wants advice on getting the maximum speed from an '887, well as
far as I can see, this has been a very useful discussion (e.g. I've never
used a canned oscillator, and may not have considered the downsides that
Olin mentioned).  There was clear advice that, if you want maximum speed,
an external RC circuit is a bad way to go.

I guess what you object to is the answer to "which pin do I use" was "look
in the data sheet".  Yes, people want to get things done, but to be able
to answer that question, you don't have to be "obsessed with technology".
Almost all of us would have to go and look that one up.  Why should I (for
example) take the time to look it up on behalf of the OP, and present the
complete answer on a platter?  What is wrong with saying "here's where you
find that info, go get it"?  It's not asking them to find anything obscure
at all - it's right there, in the section on oscillators, with a simple
diagram showing a resistor and a capacitor.

Are you really saying that, if you "just want to get things done", you
shouldn't ever be expected to look anything up for yourself?


David Meiklejohn
http://www.gooligum.com.au

2008\06\17@200520 by James Nick Sears

flavicon
face
On Tue, Jun 17, 2008 at 7:47 PM, David Meiklejohn <EraseMEdavidspamspamspamBeGonegooligum.com.au> wrote:
> James Nick Sears wrote:
> I guess what you object to is the answer to "which pin do I use" was "look
> in the data sheet".  Yes, people want to get things done, but to be able
> to answer that question, you don't have to be "obsessed with technology".
> Almost all of us would have to go and look that one up.  Why should I (for
> example) take the time to look it up on behalf of the OP, and present the
> complete answer on a platter?  What is wrong with saying "here's where you
> find that info, go get it"?  It's not asking them to find anything obscure
> at all - it's right there, in the section on oscillators, with a simple
> diagram showing a resistor and a capacitor.

There's absolutely nothing wrong with that.  What baffles me a bit is
when people clearly get angry about people asking questions that are
answerable by the datasheet and then start on a big tirade about how
people want everything handed to them.  If you are too busy to look it
up in the datasheet for them, then I would presume that you should
also be too busy to rant about them asking.

>
> Are you really saying that, if you "just want to get things done", you
> shouldn't ever be expected to look anything up for yourself?
>

Not unless you've got enough money to pay me to do it for you.

-n.

2008\06\17@202648 by olin piclist

face picon face
James Nick Sears wrote:
> If you are too busy to look it
> up in the datasheet for them, then I would presume that you should
> also be too busy to rant about them asking.

Refusing to hand people information they could have gotten from the
datasheet just by looking isn't usually about being busy.  It's a matter of
principle of not letting them get away with being lazy, of teaching them a
lesson, and of setting a standard for what's acceptable.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\17@203906 by James Nick Sears

flavicon
face
On Tue, Jun 17, 2008 at 8:26 PM, Olin Lathrop <RemoveMEolin_piclistKILLspamspamembedinc.com> wrote:
> James Nick Sears wrote:
>> If you are too busy to look it
>> up in the datasheet for them, then I would presume that you should
>> also be too busy to rant about them asking.
>
> Refusing to hand people information they could have gotten from the
> datasheet just by looking isn't usually about being busy.  It's a matter of
> principle of not letting them get away with being lazy, of teaching them a
> lesson, and of setting a standard for what's acceptable.

Fair enough.  I'm sure on a different day I'd be more or less annoyed
by the same.  My point was primarily to try to bring up the point that
some people don't necessarily want to learn the ins and outs of uC,
but still want to get something done with a microcontroller, and that
there may be a somewhat different standard of what's acceptable in
that case.  But I've made that point at least a few times today and
enough is enough.

Peace?

-n.


>
>
> ********************************************************************
> Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
> (978) 742-9014.  Gold level PIC consultants since 2000.
> -

2008\06\17@210458 by Jinx

face picon face
> some people don't necessarily want to learn the ins and outs of
> uC, but still want to get something done with a microcontroller

James, I appreciate that, but there's nevertheless a big difference
between "where is that in the datasheet ?" and "I've read it and
don't understand / need help with this bit". Of course it's possible
to write code for a PIC without ever powering one up, but if you
want that code running IN a PIC you have to know how to do
that part too, and that's when you have to have a little patience
and RTFM

2008\06\17@224957 by Gaston Gagnon

face
flavicon
face

>
> Olin Lathrop wrote:
>> Refusing to hand people information they could have gotten from the
>> datasheet just by looking isn't usually about being busy.  It's a
>> matter of
>> principle of not letting them get away with being lazy, of teaching
>> them a
>> lesson, and of setting a standard for what's acceptable.
>>
>>  
>
Please Bob, would it be within the realm of your power to put EVERYBODY,
ALL POSTERS on moderation under Olin?

The objective of this would be to prevent substandard requests and even
substandard solutions from incompetent posters from reaching  the list.  
This should result in a peaceful, quiet, civilized list with acceptable
standards. That says nothing of the traffic cut and the time saved by
2000 piclisters reading those stupid questions from "lazy" people.
Another approach, even better, I think, would be to put in the list's
rules that everybody should read everything published to date and find
the optimum solution to their problems before even thinking of posting.
Wouldn't that be neat? The piclisters would be so busy reading manuals
that only Olin would be available to ask himself quality questions and
of course, also the only one available to provide adequate answers
leaving the other experts in awe and admiration.

Lets fix that broken list once and for all.

Gaston (half joking) :-\

2008\06\18@014722 by Wouter van Ooijen

face picon face
Tomás Ó hÉilidhe wrote:
>
> Wouter van Ooijen wrote:
>> The proper answer is (if you want reliable maximum speed): none, use a
>> resonator, crystal, or canned oscillator instead. And read the posts
>> about those. IMHO a 3-leg resonator (with build-in caps) is the most
>> hassle-free, but also the least accurate (but for most purposes,
>> including serail asynch comm, accurate enough).
>
> I don't need precision. It can be 18 MHz, it can be 21 MHz.

No it can't: 21 MHz is outside the specs! That's one reason an external
RC oscillator at 20 Mhz is not practical.

> Can you please suggest a resonator part number that I can look up?

I sell the thingies (through hole), but the part number is either my
product number or the manufacturers number (and I doubt you will find
exactly the same part anywhere else).

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\06\18@035238 by David Meiklejohn

face
flavicon
face
Wouter van Ooijen wrote:
> Tomás Ó hÉilidhe wrote:
>>
>> I don't need precision. It can be 18 MHz, it can be 21 MHz.
>
> No it can't: 21 MHz is outside the specs! That's one reason an external
> RC oscillator at 20 Mhz is not practical.

This makes me wonder (please note - I am NOT advocating this, I don't want
to do it - it's just curiosity), has anyone here tried "overclocking" any
PICs?  How did it go?  What speeds did you achieve?

Again - just to be really clear before I'm jumped on, I'd never suggest
doing it in any application you want reliability from - and that includes
hobbies.  I'm just wondering how much "safety factor" is in the specs, in
practice.


David Meiklejohn
http://www.gooligum.com.au

2008\06\18@041014 by Alan B. Pearce

face picon face
> How is that substantially easier than a crystal and two caps or a single
> ceramic resonator with internal caps?  Those canned oscillators are bulky,
> expensive, use a lot of power, and glitch the supply.  At the minimum you
> need two parts, the oscillator and the bypass cap.

I would hardly describe these as taking up more square inches (or mm if you
prefer) than a crystal and a couple of capacitors.

http://onecall.farnell.com/1276643/passives/product.us0?sku=rakon-cfps-39ib-20-0mhz&_requestid=42065

The only problem I see is the 20mA draw.

2008\06\18@041114 by Richard Prosser

picon face
David,

I did try running an old 16C73A at a higher clock speed than spec'd. I
can't remember the exact speed I got to but it was a lot faster than
spec (5 times ??) and the first thing to apparently cause a problem
was the UART. The ADC was still working although was likely to be
somewhat inaccurate. The chip was starting to get quite warm also
IIRC. But the old "C' series PICs were quite tough - on another
occasion I connected one up backwards (accidentaly) and it was 'finger
burning hot' before I discovered the problem. And it still worked
afterwards.

RP

2008/6/18 David Meiklejohn <davidSTOPspamspamspam_OUTgooligum.com.au>:
{Quote hidden}

>

2008\06\18@042322 by Wouter van Ooijen

face picon face
> This makes me wonder (please note - I am NOT advocating this, I don't want
> to do it - it's just curiosity), has anyone here tried "overclocking" any
> PICs?  How did it go?  What speeds did you achieve?

IIRC there was a design (for a video game) that used a 10 MHz chip on 12
or 16 MHz (google to find the real figure). It seemed to work, and IIRC
it seemed not to work at 18 MHz.

But keep in mind that no-one here can do the kind of testing Microchip
can do, so there is absolutely no way to say that a chip will reliably
work at let's say 22 MHz instead of 20 MHz. The only thing one can say
is that this one particular chip on my desk *appears* to work in this
particular case (power, filtering, clock, application, temperature,
phase of the moon, proximity of dead fish, etc).

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\06\18@042646 by peter green

flavicon
face
David Meiklejohn wrote:
> Wouter van Ooijen wrote:
>  
>> Tomás Ó hÉilidhe wrote:
>>    
>>> I don't need precision. It can be 18 MHz, it can be 21 MHz.
>>>      
>> No it can't: 21 MHz is outside the specs! That's one reason an external
>> RC oscillator at 20 Mhz is not practical.
>>    
>
> This makes me wonder (please note - I am NOT advocating this, I don't want
> to do it - it's just curiosity), has anyone here tried "overclocking" any
> PICs?  How did it go?  What speeds did you achieve?
I have heared of someone accidently overclocking a 18F pic to seventy
something megahertz by selecting the wrong oscilator mode (hspll instead
of hs). apparently the thing ran at least to some extent.

2008\06\18@043904 by Xiaofan Chen

face picon face
On 6/18/08, David Meiklejohn <spamBeGonedavidSTOPspamspamEraseMEgooligum.com.au> wrote:
> This makes me wonder (please note - I am NOT advocating this, I don't want
> to do it - it's just curiosity), has anyone here tried "overclocking" any
> PICs?  How did it go?  What speeds did you achieve?
>
> Again - just to be really clear before I'm jumped on, I'd never suggest
> doing it in any application you want reliability from - and that includes
> hobbies.  I'm just wondering how much "safety factor" is in the specs, in
> practice.
>

For example, we can use the 3V3 18LF4550 on the Explorer 16 demo
board running at 48MHz even though the spec says that you can run
the core only at up to 20MHz at 3V3 (for USB only 16MHz).

http://forum.microchip.com/tm.aspx?m=338210

I would not use it for a real project. But it runs fine at home.

Regards,
Xiaofan

2008\06\18@050424 by Jinx

face picon face
> I have heared of someone accidently overclocking a 18F pic to
> seventy something megahertz by selecting the wrong oscilator
> mode (hspll instead of hs). apparently the thing ran at least to some
> extent

Could have been me, I've mentioned it before. I intended to use
PLL with a 9.8304MHz to get 39.3216MHz crystal but put an
18.432MHz in by mistake. AFAICT this 18F452 ran OK at
78.6432MHz and several tests seemed to proved that it was
executing instructions properly. As I've found with other trials
though, some modules may not be happy when over-clocked

Presumably it is still the case that Microchip sort micros by speed
group, not speed per se. So a -04 may run at 9.999MHz, but it
falls short of the 10MHz to put it in the -10 group. In that example
the PIC could be run at almost 10MHz and genuinely be within its
electrical and performances capabilities. It is in fact a -9.999 PIC.
However, you don't know its actual tested limit, and each batch is
likely to be different

IMVHO it's fair to say though that "conservative" over-clocking
(eg nudging a -04 to 4.096) might probably work reliably, as the
gaps between -04 / -10 and -10 / -20 are quite wide. And for
Microchip to absolutely guarantee -04 operation at 4MHz under
all spec conditions, you might assume that they allow a good
margin of error. Although only Microchip know what that margin
really is, and especially when costs get squeezed

Of course you have no complaint to take to Microchip if your
device fails to perform properly at greater than spec speed

2008\06\18@063605 by olin piclist

face picon face
David Meiklejohn wrote:
> I'm just wondering how much "safety factor"
> is in the specs, in practice.

No amount of your own testing can give you a reliable answer other than
"none".


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\18@063929 by Richard Seriani, Sr.

picon face
Here is just one of many overclocking tests:
http://www.sparkfun.com/commerce/present.php?p=Overclocking%20a%20PIC

Lots more on Google search of pic overclocking.


{Original Message removed}

2008\06\18@064213 by olin piclist

face picon face
Richard Prosser wrote:
> and the first thing to apparently cause a problem

Exactly.  You don't know how many other problems were there that your test
didn't happen to uncover.  Did you test all possible instructions with all
possible data values?  Did you try all possible sequences of data and
address bus patterns?  Did all possible data paths get tested at all worst
cases?

And even if you were somehow able to test your one PIC exhaustively, you
still can't assume that those results apply to the next PIC.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\06\18@065911 by Jake Anderson

flavicon
face
Olin Lathrop wrote:
> Richard Prosser wrote:
>  
>> and the first thing to apparently cause a problem
>>    
>
> Exactly.  You don't know how many other problems were there that your test
> didn't happen to uncover.  Did you test all possible instructions with all
> possible data values?  Did you try all possible sequences of data and
> address bus patterns?  Did all possible data paths get tested at all worst
> cases?
>
> And even if you were somehow able to test your one PIC exhaustively, you
> still can't assume that those results apply to the next PIC.
>
>
> ********************************************************************
> Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
> (978) 742-9014.  Gold level PIC consultants since 2000.
>  

Does microchip do all that for every pic?

2008\06\18@070500 by David Meiklejohn

face
flavicon
face
Olin Lathrop wrote:
>
> David Meiklejohn wrote:
> > I'm just wondering how much "safety factor"
> > is in the specs, in practice.
>
> No amount of your own testing can give you a reliable answer other than
> "none".

Very true.  Like I said, it's not something I'd do myself - I'm a firm
believer in operating within published specs.  But, there's also such a
thing as idle curiosity, the search for knowledge with no practical value,
and all that!  :-)


David



2008\06\18@071641 by M. Adam Davis

face picon face
On Wed, Jun 18, 2008 at 6:38 AM, Olin Lathrop <KILLspamolin_piclistspamBeGonespamembedinc.com> wrote:
> David Meiklejohn wrote:
>> I'm just wondering how much "safety factor"
>> is in the specs, in practice.
>
> No amount of your own testing can give you a reliable answer other than
> "none".

And yet some companies build on the idea that they can.  ZWorld, for
instance, started off by selling SBCs with overclocked Z80s in the
1980's.  They went through fairly exhaustive testing on many chips to
verify that it would work generally, and then they tested each board
that came off the line fairly extensively, and had a liberal return
policy.

IIRC they never had a failure that was due to overclocking, but the
overclocking was less than 2x the rated max.

As far as current chip manufacturers, though, even though they
fabricate the chip they still have to test them to determine,
statistically, what the maximum speed is - Microchip doesn't produce
chips that stop working at 4, 20, 32, 40, etc MHz, they produce
batches of chips that will work 99.995% of the time below their rated
max.

If you want to run your chip at 2x, 5x, or 10x the rated max you'll
find that a certain percentage of chips will run at those speeds.

Further, the silicon isn't different for the -20 and the -4, and
depending on market demand they may be marking -20 capable chips as
-4.

Lastly, I haven't yet met an engineer or company that doesn't
under-rate their parts to some degree to add some margin.

So, I disagree - one can do enough testing to guarantee that a
statistical percentage of your product is going to be fine over the
product's normal life.  The manufacturer of the product itself does
this.

The big caveat here is that now you are working outside the
manufacturer's recommended maximums you cannot rely on them for
support and you must start to deal with issues that are very silicon
specific - which you may not have much information for.

It's a lot of extra work, and given the wide range of available,
cheap, fast 32 bit processors I have to wonder if there's ever any
financial justification for overclocking a PIC.

-Adam

--
EARTH DAY 2008
Tuesday April 22
Save Money * Save Oil * Save Lives * Save the Planet
http://www.driveslowly.org

2008\06\18@082455 by Spehro Pefhany

picon face
At 07:04 AM 6/18/2008, you wrote:
>Olin Lathrop wrote:
> >
> > David Meiklejohn wrote:
> > > I'm just wondering how much "safety factor"
> > > is in the specs, in practice.
> >
> > No amount of your own testing can give you a reliable answer other than
> > "none".
>
>Very true.  Like I said, it's not something I'd do myself - I'm a firm
>believer in operating within published specs.  But, there's also such a
>thing as idle curiosity, the search for knowledge with no practical value,
>and all that!  :-)

Given the assumption that (based on solid-state physics and CMOS design)
there's a predictable change of maximum fosc with
temperature, one could probably quite safely get a small increase in fosc
above rated maximum by restricting the Tj range. Since Microchip do not
test every chip at temperature extremes, they would have to test the chips
to a tighter tolerance at room temperature to guarantee most chips will meet
specs at the extremes. I'd feel quite comfortable doing this in a production
situation, for example to get 20.48MHz = 5 * 4096MHz operation on a 20MHz
rated part by using a part rated and tested for (say) 125C°C at no more
than 60°C Tj (and thus perhaps 50°C Ta). This might be important to generate
a low-jitter high speed clock at a specific "odd" frequency, for example.

Of course even so there's a small amount of risk in this, and the safe
improvement is not a very high percentage-- and there are usually better
choices that don't require this sort of thing, so I don't recommend it.

>Best regards,

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



2008\06\18@085248 by Dario Greggio

face picon face
Jan-Erik Soderholm wrote:

>>If you haven't read
>>the datasheet cover to cover, be damned.
>
> I do not know about "cover to cover", the
> actual part about the RC oscillator would
> probably be enough. I do not think it's even

yep, same thought here :)

--
Ciao, Dario

2008\06\18@085401 by Tomás Ó hÉilidhe

picon face


David Meiklejohn wrote:
> This makes me wonder (please note - I am NOT advocating this, I don't want
> to do it - it's just curiosity), has anyone here tried "overclocking" any
> PICs?  How did it go?  What speeds did you achieve?

My guess is you could probably get WAY more than 20 MHz out of it.

PIC datasheets are extremely safe, from my experience with them. For a
college project, I powered LED's from PIC pins and didn't use
current-limiting resistors. I had seven of these on at once. Not only
was I taking more current from a pin that I should have, but I also took
too much current overall for the entire chip. But the project board
worked perfectly, I left it on overnight for three nights in a row
without a glitch.

(Of course, three nights isn't equal to 30 years, but it shows what
these chips can do)

2008\06\18@090250 by Dario Greggio

face picon face
David Meiklejohn wrote:

>
> This makes me wonder (please note - I am NOT advocating this, I don't want
> to do it - it's just curiosity), has anyone here tried "overclocking" any
> PICs?  How did it go?  What speeds did you achieve?
>

I tried 20-25% and is perfectly ok on 16F and 18F parts.

--
Ciao, Dario -- ADPM Synthesis sas -- http://www.adpm.tk

2008\06\18@090753 by Dario Greggio

face picon face
Jinx wrote:

>>I have heared of someone accidently overclocking a 18F pic to
>>seventy something megahertz by selecting the wrong oscilator
>>mode (hspll instead of hs). apparently the thing ran at least to some
>>extent
>
> Could have been me, I've mentioned it before. I intended to use


Yeah, or me too :) 48MHz in USB and/or @3.3V

Funny, at the moment, since 1 hour, the forum can't be accessed from here.


I'd add to the thread, that possibly "giving more voltage to the PIC"
might help overclocking more.. of course with added risks!

--
Ciao, Dario -- ADPM Synthesis sas -- http://www.adpm.tk

2008\06\18@095908 by Nicola Perotto

picon face


Tomás Ó hÉilidhe wrote:
> PIC datasheets are extremely safe, from my experience with them.
uah uah uah


(sorry i can't resist ;-)

2008\06\18@100840 by Spehro Pefhany

picon face
Quoting Dario Greggio <@spam@adpm.to@spam@spamspam_OUTinwind.it>:

> Jan-Erik Soderholm wrote:
>
>>> If you haven't read
>>> the datasheet cover to cover, be damned.
>>
>> I do not know about "cover to cover", the
>> actual part about the RC oscillator would
>> probably be enough. I do not think it's even
>
> yep, same thought here :)

But how would you find that exact section? Is there some kind of table  
of contents or "Index" feature? Or is it possible to "search" for the  
information? Anyone know?


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

2008\06\18@102311 by Stephen D. Barnes

flavicon
face
Spehro Pefhany wrote:
> But how would you find that exact section? Is there some kind of table  
> of contents or "Index" feature? Or is it possible to "search" for the  
> information? Anyone know?
>
>
> Best regards,
> Spehro Pefhany
>  
Since the Microchip data sheets are in PDF format, IF you are using
Adobe Acrobat Reader, just use the search feature. I do this all the time.

--
Regards,
Stephen D. Barnes

2008\06\18@102701 by Dario Greggio

face picon face
Spehro Pefhany wrote:
> But how would you find that exact section? Is there some kind of table  
> of contents or "Index" feature? Or is it possible to "search" for the  
> information? Anyone know?

Hmmm, the Index at the left is usually good enough: it is divided into
Sections almost self-explicative.

Search through the text works - but not always, at least Acrobat seems
to go crazy about it . I 'm using FoxIt Reader and looks a bit better
(besides MUCH faster).

When you can't find something, the "Special Features" is worth a look :)
, or Google or the Forum.

--
Ciao, Dario

2008\06\18@103856 by Wouter van Ooijen

face picon face
Spehro Pefhany wrote:
> Quoting Dario Greggio <.....adpm.tospam_OUTspaminwind.it>:
>
>> Jan-Erik Soderholm wrote:
>>
>>>> If you haven't read
>>>> the datasheet cover to cover, be damned.
>>> I do not know about "cover to cover", the
>>> actual part about the RC oscillator would
>>> probably be enough. I do not think it's even
>> yep, same thought here :)
>
> But how would you find that exact section? Is there some kind of table  
> of contents or "Index" feature? Or is it possible to "search" for the  
> information? Anyone know?

The datasheets do have an Index (at least some have TOC), and (when
opened in acrobat) the page number in the index are hyperlinks.
"oscillator" for instance is an index entry.

Oscillator settings and most (all?) other things that have to do with
the fuses are found in the "special features of the CPU section".

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\06\18@104038 by Alan B. Pearce

face picon face
>But how would you find that exact section? Is there some kind of table
>of contents or "Index" feature? Or is it possible to "search" for the
>information? Anyone know?

I don't know about the paper copy, but the PDF ones certainly have an index
available, and are searchable.

2008\06\18@104433 by Jan-Erik Soderholm

face picon face
Spehro Pefhany wrote:
> Quoting Dario Greggio <TakeThisOuTadpm.to.....spamTakeThisOuTinwind.it>:
>
>> Jan-Erik Soderholm wrote:
>>
>>>> If you haven't read
>>>> the datasheet cover to cover, be damned.
>>> I do not know about "cover to cover", the
>>> actual part about the RC oscillator would
>>> probably be enough. I do not think it's even
>> yep, same thought here :)
>
> But how would you find that exact section? Is there some kind of table  
> of contents or "Index" feature? Or is it possible to "search" for the  
> information? Anyone know?

We are talking about "oscillator" issues, right ?
So the chapter called "4.0 OSCILLATOR MODULE..."
might be a good start, not ?

Then browsning 4-5 pages forward in that chapter, there
is a section called "4.4.4 EXTERNAL RC MODES". Well now
we are realy close, aren't we ?

Actualy, you do not have to open the 4.0 chapter as such,
it's enough to just look in the "bookmark" section
in the PDF where you can "open" the chapters and drill-down.

Jan-Erik.

2008\06\18@120128 by Richard Seriani, Sr.

picon face
Tomás,

You stated, "...I powered LED's from PIC pins and didn't use
current-limiting resistors. I had seven of these on at once. Not only was I
taking more current from a pin that I should have, but I also took too much
current overall for the entire chip."

Questions:
1. Which PIC?
2. At what speed and supply voltage were you running it?
3. Were you sinking or sourcing current from/to the LEDs?
4. Which pins were you using (multiple ports or all on one port)?
5. Which LEDs (current/voltage specs)?
6. How were the LEDs connected?
7. Were there only seven total LEDs?
8. If not, were the same seven LED's always on during the test?
9. If all LEDs were always on, was there no change taking place in the
program? In other words, was it always at the same line once the LEDs were
turned on?
10. If no change, how did you determine there wasn't a glitch?

"taking more..." and "...took too much..." don't mean much without more
specific information. Please explain.

Thanks,
Richard


{Original Message removed}

2008\06\18@130530 by Tomás Ó hÉilidhe

picon face


Richard Seriani, Sr. wrote:
> Tomás,
>
> You stated, "...I powered LED's from PIC pins and didn't use
> current-limiting resistors. I had seven of these on at once. Not only was I
> taking more current from a pin that I should have, but I also took too much
> current overall for the entire chip."
>
> Questions:
> 1. Which PIC?

16F684

> 2. At what speed and supply voltage were you running it?
>  

8 MHz internal oscillator. 5 volts.

> 3. Were you sinking or sourcing current from/to the LEDs?
>  

Both. I had a display multiplex design. The first 8 steps of the
multiplexer would source current, the second 8 steps would sink. Each
step was kept lit for 2 milliseconds.

> 4. Which pins were you using (multiple ports or all on one port)?
>  

It has 14 pins available for I/O and I used every single one of them.
PORTC was used for driving the LED's.

> 5. Which LEDs (current/voltage specs)?
>  

Two-pin bi-colour LED's. They have about 2.2 volts across them when the
current is 25 mA.

The pins however were driving transistors, BC337 and BC327. I didn't put
in a base transistor.

> 6. How were the LEDs connected?
>  

Here's the current path:

* 5 V on microcontroller chip
to
* Diode with .7 V across it
to
* Ground

> 7. Were there only seven total LEDs?
>  

50 bi-colour LED's altogether arranged in a matrix of 7 x 7 (which one
extra).

The multiplexer only lit one column at a time.

> 8. If not, were the same seven LED's always on during the test?
>  

I could turn every LED of the matrix on in code, and the display
multiplexer would light them all accordingly without a hitch.

> 9. If all LEDs were always on, was there no change taking place in the
> program? In other words, was it always at the same line once the LEDs were
> turned on?
>  

Multiplexer didn't screw up at all, not even once over 3 days of being
on constantly.

> 10. If no change, how did you determine there wasn't a glitch?
>  

Display had the same pattern on it every morning.
> "taking more..." and "...took too much..." don't mean much without more
> specific information. Please explain.

Well if you take a uC pin that has 5 volts on it, put it into a diode,
and then put the diode directly to ground, then it will at least  try to
draw a hell of a lot more than 25 mA. My guess is that it even caused a
voltage drop within the uC pin itself.

2008\06\18@155329 by James Nick Sears

flavicon
face
I've seen dozens, if not hundreds, of PIC projects done in just this
way, with no limiting resistors and have never seen a PIC smoke or
otherwise seriously malfunction.  Those output drivers are indeed
pretty tough.  Or at least have enough output resistance to save
themselves in this scenario.

It does guarantee some really nasty RMW issues though if the Vf of the
LED is less than the PIC's Vih at If=~20mA.

-n.



On Wed, Jun 18, 2008 at 1:05 PM, Tomás Ó hÉilidhe <TakeThisOuTtoeKILLspamspamspamlavabit.com> wrote:
{Quote hidden}

>

2008\06\18@203324 by M. Adam Davis

face picon face
CMOS output drivers of the type the PIC has are reasonably
self-limiting - as the current draw goes up so does the resistance.

However, they are not rated for high current draw and you will see
failures if you put this in production.

Other odd things happen at high currents as well - you'll see a lot
more inductive coupling in other areas of the silicon when they
switch, which may cause occasional glitches - you wouldn't necessarily
notice in a given project, but in a critical application it might ruin
your day.

-Adam

On Wed, Jun 18, 2008 at 3:53 PM, James Nick Sears <.....listsspamRemoveMEjamesnsears.com> wrote:
{Quote hidden}

>> -

2008\06\18@222819 by William \Chops\ Westfield

face picon face

On Jun 18, 2008, at 3:39 AM, Richard Seriani, Sr. wrote:

> has anyone here tried "overclocking" any
> PICs?  How did it go?  What speeds did you achieve?

It might be worth adding to the discussion that it is apparently  
pretty common for the crystal oscillator to be one of the first  
pieces that stops working as clock speeds go up.  IIRC, there are  
even some chips (scenix?) that are spec'ed to run at their top speed  
ONLY when driven by an external oscillator.  So if you're planning to  
try over-clocking a PIC, you might weight your oscillator selection  
toward the external canned oscillators...

BillW

2008\06\19@051815 by Michael Rigby-Jones

picon face


> -----Original Message-----
> From: spamBeGonepiclist-bounces@spam@spamspam_OUTmit.edu [TakeThisOuTpiclist-bouncesspamspammit.edu] On
Behalf
{Quote hidden}

I've also seen this plenty of times, again with smoke.  I wonder if this
is due to laziness or just ignorance of LED characteristics (i.e. 'it's
just a small light bulb')?

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2008\06\19@060112 by James Nick Sears

flavicon
face
On Thu, Jun 19, 2008 at 5:17 AM, Michael Rigby-Jones
<Michael.Rigby-JonesEraseMEspambookham.com> wrote:
>
>
>> {Original Message removed}

2008\06\19@130314 by Michael Rigby-Jones

picon face


> -----Original Message-----
> From: RemoveMEpiclist-bouncesEraseMEspamspam_OUTmit.edu [@spam@piclist-bouncesRemoveMEspamEraseMEmit.edu] On
Behalf
{Quote hidden}

this
> >> way, with no limiting resistors and have never seen a PIC smoke or
> >> otherwise seriously malfunction.  Those output drivers are indeed
> >> pretty tough.  Or at least have enough output resistance to save
> >> themselves in this scenario.
> >
> > I've also seen this plenty of times, again with smoke.  I wonder if
this
> > is due to laziness or just ignorance of LED characteristics (i.e.
'it's
> > just a small light bulb')?
>
> If you're smoking lots of PICs by driving LEDs directly from the pins
> then I'd advise you to check your balance at the First Bank of Karma.
>
> Probably the people doing it are simply inferior humans.  I bet that's
> also why the universe feels sorry for them and it works for them so
> much of the time.

Heh, I meant to say "without".  PIC's are pretty tough little things
really.

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================


'[PIC] Max speed from 887'
2008\07\01@195258 by Forrest W. Christian
flavicon
face
Bob Blick wrote:
> I doubt you'll find anyone here who has used RC mode on an '887. Use a
> crystal or approved ceramic resonator(I posted resonator tests a while
> back, some don't work so well at 20MHz).
Plus, RC mode is rated 0-4Mhz (see table 17-1 in the datasheet).   The
original poster wanted max speed, so if your options are the 8Mhz
internal oscillator or a 4Mhz RC oscillator, I think the internal
oscillator would win.

Like others have mentioned, the PLL's in the newer PIC's are really
nice... the 18F24K20 for instance has a 16Mhz intosc, and a 4xPLL,
giving you a 64Mhz clock without an external crystal.

-forrest

More... (looser matching)
- Last day of these posts
- In 2008 , 2009 only
- Today
- New search...