Searching \ for '[PIC] Running a PIC near it's maximum rated freque' 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: 'Running a PIC near it's maximum rated freque'.

Exact match. Not showing close matches.
PICList Thread
'[PIC] Running a PIC near it's maximum rated freque'
2010\07\21@181456 by V G

picon face
Hey all,

I'm planning to use an external crystal oscillator (not a crystal as a
discrete component, but a crystal oscillator IC) for my
dsPIC33FJ128GP802. The dsPIC is rated to run at a maximum of 40 MIPS
on 3 V - 3.3 V. Would running it with a 40 MHz crystal oscillator be
safe? As in, would it be reliable to run the dsPIC at it's maximum
rated frequency?

2010\07\21@193424 by Oli Glaser

flavicon
face


--------------------------------------------------
From: "V G" <spam_OUTx.solarwind.xTakeThisOuTspamgmail.com>
Sent: Wednesday, July 21, 2010 11:14 PM
To: "PICLIST" <.....piclistKILLspamspam@spam@mit.edu>
Subject: [PIC] Running a PIC near it's maximum rated frequency

> Hey all,
>
> I'm planning to use an external crystal oscillator (not a crystal as a
> discrete component, but a crystal oscillator IC) for my
> dsPIC33FJ128GP802. The dsPIC is rated to run at a maximum of 40 MIPS
> on 3 V - 3.3 V. Would running it with a 40 MHz crystal oscillator be
> safe? As in, would it be reliable to run the dsPIC at it's maximum
> rated frequency?

According to the data sheet it can use an external clock up to 64MHz, and a
crystal up to 40MHz, so I would guess it should be fine as long as
reasonable precautions are taken with regard to trace length, capacitance
etc.



2010\07\21@194501 by V G

picon face
On Wed, Jul 21, 2010 at 7:34 PM, Oli Glaser <oli.glaserspamKILLspamtalktalk.net> wrote:
> According to the data sheet it can use an external clock up to 64MHz, and a
> crystal up to 40MHz, so I would guess it should be fine as long as
> reasonable precautions are taken with regard to trace length, capacitance
> etc.

But it says maximum 40 MIPS. Is there some way to get more than 40 MIPS?

2010\07\21@194744 by V G

picon face
On Wed, Jul 21, 2010 at 7:44 PM, V G <.....x.solarwind.xKILLspamspam.....gmail.com> wrote:
> On Wed, Jul 21, 2010 at 7:34 PM, Oli Glaser <EraseMEoli.glaserspam_OUTspamTakeThisOuTtalktalk.net> wrote:
>> According to the data sheet it can use an external clock up to 64MHz, and a
>> crystal up to 40MHz, so I would guess it should be fine as long as
>> reasonable precautions are taken with regard to trace length, capacitance
>> etc.
>
> But it says maximum 40 MIPS. Is there some way to get more than 40 MIPS?

Oh! I just realized that the instruction clock is half of the
oscillator frequency. So I can still go with a 40 MHz crystal
oscillator and double it via PLL to 80 MHz to get the full 40 MIPS out
of it?

2010\07\21@195700 by V G

picon face
On Wed, Jul 21, 2010 at 7:47 PM, V G <x.solarwind.xspamspam_OUTgmail.com> wrote:
> On Wed, Jul 21, 2010 at 7:44 PM, V G <@spam@x.solarwind.xKILLspamspamgmail.com> wrote:
>> On Wed, Jul 21, 2010 at 7:34 PM, Oli Glaser <KILLspamoli.glaserKILLspamspamtalktalk.net> wrote:
>>> According to the data sheet it can use an external clock up to 64MHz, and a
>>> crystal up to 40MHz, so I would guess it should be fine as long as
>>> reasonable precautions are taken with regard to trace length, capacitance
>>> etc.
>>
>> But it says maximum 40 MIPS. Is there some way to get more than 40 MIPS?
>
> Oh! I just realized that the instruction clock is half of the
> oscillator frequency. So I can still go with a 40 MHz crystal
> oscillator and double it via PLL to 80 MHz to get the full 40 MIPS out
> of it?


Also, the data sheet doesn't say anything about a max 64 MHz clock
source for the main oscillator. It only says that for the axillary
clock source for peripherals. So I'm guessing I could use an 80 MHz
crystal oscillator if I wanted.

2010\07\21@195713 by Brendan Gillatt

flavicon
face
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 22/07/2010 00:44, V G wrote:
> On Wed, Jul 21, 2010 at 7:34 PM, Oli Glaser <RemoveMEoli.glaserTakeThisOuTspamtalktalk.net> wrote:
>> According to the data sheet it can use an external clock up to 64MHz, and a
>> crystal up to 40MHz, so I would guess it should be fine as long as
>> reasonable precautions are taken with regard to trace length, capacitance
>> etc.
>
> But it says maximum 40 MIPS. Is there some way to get more than 40 MIPS?

Probably, but you have no manufacturer guarantee that it will work (i.e.
do your own tests).

Usually the limitation of flash micro controllers is the access time of
the flash memory itself. Above 30-50 mega-fetches / second, wait states
have to be used to ensure instructions are read correctly. With PICs (no
wait state support) the flash may not have enough time to fetch the next
instruction before the pipeline advances.

- --
Brendan Gillatt | GPG Key: 0xBF6A0D94
brendan {a} brendangillatt (dot) co (dot) uk
http://www.brendangillatt.co.uk
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.14 (MingW32)

iEYEARECAAYFAkxHiVUACgkQHEhZ5Ws5poG+/wCgo9/n6PnM54fOoyrwpwTM5y64
lWoAoKlgky8+Vw6SFE3jyikNth4B4U5U
=W/k5
-----END PGP SIGNATURE-----

2010\07\21@200234 by Oli Glaser

flavicon
face


--------------------------------------------------
From: "V G" <spamBeGonex.solarwind.xspamBeGonespamgmail.com>
Sent: Thursday, July 22, 2010 12:47 AM
To: "Microcontroller discussion list - Public." <TakeThisOuTpiclistEraseMEspamspam_OUTmit.edu>
Subject: Re: [PIC] Running a PIC near it's maximum rated frequency

{Quote hidden}

Yes, I think you can set the PLL to give a range of frequencies. For
instance, I think you could use a 10MHz crystal and run it at maximum MIPS
using the right settings.


2010\07\21@224154 by RussellMc

face picon face
> I'm planning to use an external crystal oscillator (not a crystal as a
> discrete component, but a crystal oscillator IC) for my
> dsPIC33FJ128GP802. The dsPIC is rated to run at a maximum of 40 MIPS
> on 3 V - 3.3 V. Would running it with a 40 MHz crystal oscillator be
> safe? As in, would it be reliable to run the dsPIC at it's maximum
> rated frequency?

The following is meant to be genuinely useful and not just pedantic.
(To some degree, also being pedantic is necessary to be sure of what
you are getting).

It's highly likely (never certain) that the data sheet lays all this
out reasonably clearly. This doesn't mean that interpreting them will
be easy or even guaranteed to be unambiguous. So, asking questions is
fine, but care is needed.

You are mixing various parameters within the same question set -
"mixed units" as it were. eg MIPS and MHz and types of clock and clock
sources etc. This is OK enough as long as you don't confuse yourself
with the mixing and as long as when you get an answer you make sure
that the person answering it is answering what you think they are.
It's easy for either or both of these not to be the case. ie it's easy
for ambiguity to lead to classic GIGO.

It is usually better to ask questions which are as unambiguous as
possible so that minimum effort is subsequently spent on clarification
and maximum confidence can be placed in answers.

Below "max" means "maximum value allowed by the data sheet for normal
operation, taking due note of relevant voltages loadings etc".

All these are to a greater or lesser extent independent:

- MIPS,
- Max allowable crystal oscillator frequency.
- Max PLL input frequency.
- Max PLL output frequency.
- Max allowed PLL output frequency at a given input frequency.
- Max allowed clock input on a given input pin.
- More

Look at your questions so far and answers received and try and assess
how they relate.

Do you have a datasheet link - it's often useful to other people to
post such with your query so they can access it easily and also so
they can establish what you think you are working with. Sometimes
claimed data sheet and device can differ.


      Russell McMahon

2010\07\22@040045 by RussellMc

face picon face
> I'm planning to use an external crystal oscillator (not a crystal as a
> discrete component, but a crystal oscillator IC) for my
> dsPIC33FJ128GP802. The dsPIC is rated to run at a maximum of 40 MIPS
> on 3 V - 3.3 V. Would running it with a 40 MHz crystal oscillator be
> safe? As in, would it be reliable to run the dsPIC at it's maximum
> rated frequency?

40 MHz is apparently much higher than needed but appears valid. needed.

eg see section 9.1.4 of
http://ww1.microchip.com/downloads/en/DeviceDoc/70292D.pdf

1. With crystal (not what you asked)
10 MHz crystal with PLLPRE=0, PLLDIV=ox1E, PLLPOST = 0
Gives: VCO input = 5 MHz, VCO output = 160 MHZ,  Fosc = 980 MHz.
MIPS = 40
QED.

2. Figure 9-22 shows requirements.
PLLPRE must be 0.8 - 8 MHz.

Table 30-16 suggests an external clock of <= 40 Mhz in ECPLL mode BUT
you can use far lower if desired.
eg as low as 0.8 x 2 = 1.6 Mhz.

SO
- 40 MHz EC meets PLL capabilities for 40 MIPS.
Highest allowed EC is probably 40 MHz.
Lowest allowed EC is 1.6 MHz.

E&OE.


R


                     Russell

2010\07\22@042818 by Oli Glaser

flavicon
face


--------------------------------------------------
From: "RussellMc" <EraseMEapptechnzspamgmail.com>
Sent: Thursday, July 22, 2010 9:00 AM
To: "Microcontroller discussion list - Public." <RemoveMEpiclistEraseMEspamEraseMEmit.edu>
Subject: Re: [PIC] Running a PIC near it's maximum rated frequency

{Quote hidden}

I'm assuming you meant Fosc = 80MHz here?


{Quote hidden}

> --

2010\07\22@045925 by RussellMc

face picon face
>> eg see section 9.1.4 of
>> ww1.microchip.com/downloads/en/DeviceDoc/70292D.pdf
>>
Russell McMahon wrote:
>> 10 MHz crystal with PLLPRE=0, PLLDIV=ox1E, PLLPOST = 0
>> Gives: VCO input = 5 MHz, VCO output = 160 MHZ,  Fosc = 980 MHz.
>> MIPS = 40
>> QED.
>
> I'm assuming you meant Fosc = 80MHz here?

Aye - double tap apparently.



       Russell

2010\07\22@072531 by Olin Lathrop

face picon face
V G wrote:
> As in, would it be reliable to run the dsPIC at it's maximum
> rated frequency?

Of course, since that would be within spec.  However, check what the maximum
oscillator input frequency is.  That is a different spec from the maximum
instruction cycle rate.


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

2010\07\22@073008 by Olin Lathrop

face picon face
V G wrote:
> But it says maximum 40 MIPS. Is there some way to get more than 40
> MIPS?

Look up "maximum" in a dictionary.

2010\07\22@073135 by Olin Lathrop

face picon face
V G wrote:
> Oh! I just realized that the instruction clock is half of the
> oscillator frequency. So I can still go with a 40 MHz crystal
> oscillator and double it via PLL to 80 MHz to get the full 40 MIPS out
> of it?

This is getting silly.  READ THE DATASHEET, particularly the whole
oscillator section and the electrical characteristics pertaining to the
oscillator.


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

2010\07\22@075456 by RussellMc

face picon face
> > Oh! I just realized that the instruction clock is half of the
> > oscillator frequency. So I can still go with a 40 MHz crystal
> > oscillator and double it via PLL to 80 MHz to get the full 40 MIPS out
> > of it?

> ... READ THE DATASHEET, ...

What he said.

And look at my my prior worked solution that suggests you may
successfully use any external clock input from 1.6 MHz to 40 Mhz,
given suitable parameter values.


      R.

2010\07\22@080213 by RussellMc

face picon face
> Also, the data sheet doesn't say anything about a max 64 MHz clock
> source for the main oscillator.

It doesn't.
But if you read my prior post (after this one of yours that I'm
replying to here) you'll see that it DOES say something about a 40 MHz
limit for the main oscillator.

> It only says that for the axillary
> clock source for peripherals. So I'm guessing I could use an 80 MHz
> crystal oscillator if I wanted.

Are you trolling for a rise from Olin?
Seems to be working :-)

ie reading about what it says about the specific signal of interest
"is wise" [tm].
Find the signal name (here "EC" and also "external clock") and search
the data sheet for numeric values.
About 99.14159265358979% of all data sheets will specify such things.
Some few fail to. This one doesn't (fail to).



       Russell

2010\07\22@080213 by RussellMc

face picon face
> Also, the data sheet doesn't say anything about a max 64 MHz clock
> source for the main oscillator.

It doesn't.
But if you read my prior post (after this one of yours that I'm
replying to here) you'll see that it DOES say something about a 40 MHz
limit for the main oscillator.

> It only says that for the axillary
> clock source for peripherals. So I'm guessing I could use an 80 MHz
> crystal oscillator if I wanted.

Are you trolling for a rise from Olin?
Seems to be working :-)

ie reading about what it says about the specific signal of interest
"is wise" [tm].
Find the signal name (here "EC" and also "external clock") and search
the data sheet for numeric values.
About 99.14159265358979% of all data sheets will specify such things.
Some few fail to. This one doesn't (fail to).



       Russell

More... (looser matching)
- Last day of these posts
- In 2010 , 2011 only
- Today
- New search...