Crystal Cload and ESR for microcontroller...

[Franco]

Hi everyone:

I was wondering if the external capacitors for matching Cload are
"always" neccesary when attaching a crystal to a microcontroller? If
the internal capacitance of the microcontroller oscillator is the same
as the specified in the datasheet of the crystal, then I don't see
reasons for the external capacitors (those two ones connected between
the the leads of the crystal and ground).

The other question is about the ESR indicated in the crystal
datasheet. If this ESR indicated in the datasheet of the crystal is
slightly higher than the one indicated as maximum for the crystal in
the datasheet of the microcontroller, how could it affect the
frequency?

Is there any way to correct that ESR of the crystal with an external
passive component (resistor or something)?

Thanks!

 

[sycochkn]

Hi everyone:

I was wondering if the external capacitors for matching Cload are
"always" neccesary when attaching a crystal to a microcontroller? If
the internal capacitance of the microcontroller oscillator is the same
as the specified in the datasheet of the crystal, then I don't see
reasons for the external capacitors (those two ones connected between
the the leads of the crystal and ground).

The other question is about the ESR indicated in the crystal
datasheet. If this ESR indicated in the datasheet of the crystal is
slightly higher than the one indicated as maximum for the crystal in
the datasheet of the microcontroller, how could it affect the
frequency?

Is there any way to correct that ESR of the crystal with an external
passive component (resistor or something)?

Thanks!

How accurate do you need the frequency to be? If you need more accuracy then
an external oven controlled crystal oscillator.

Bob

 

[Tim Wescott]

Hi everyone:

I was wondering if the external capacitors for matching Cload are
"always" neccesary when attaching a crystal to a microcontroller? If the
internal capacitance of the microcontroller oscillator is the same as
the specified in the datasheet of the crystal, then I don't see reasons
for the external capacitors (those two ones connected between the the
leads of the crystal and ground).

A microprocessor pin that has parasitic capacitance that exceeds the
crystal load capacitance would be weird. What load capacitance and
microprocessor are you using?

Depending on where the capacitance is things may not work -- you need
capacitors to ground at each oscillator terminals to properly couple the
circuit. Just putting a capacitor in parallel with the crystal will
bring the resonant frequency to the correct value, but won't allow the
oscillator to work.

The other question is about the ESR indicated in the crystal datasheet.
If this ESR indicated in the datasheet of the crystal is slightly higher
than the one indicated as maximum for the crystal in the datasheet of
the microcontroller, how could it affect the frequency?

It could bring it all the way to zero, on a bad day. The ESR of the
crystal is a factor of the overall loop gain; if the crystal is too lossy
then the oscillator may fail to start.

In general resistance in circuits like this bring the resonant frequency
down.

Is there any way to correct that ESR of the crystal with an external
passive component (resistor or something)?

No. A checkbook helps in getting better components, but it's not a
passive component when used in this manner.

--
Tim Wescott
Control systems and communications consulting
http://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html

 

[Martin Riddle]

Microchip has a good appnote worth reading.
http://ww1.microchip.com/downloads/en/AppNotes/00826a.pdf

Cheers

 

[Franco]

A microprocessor pin that has parasitic capacitance that exceeds the
crystal load capacitance would be weird. What load capacitance and
microprocessor are you using?

Depending on where the capacitance is things may not work -- you need
capacitors to ground at each oscillator terminals to properly couple the
circuit. Just putting a capacitor in parallel with the crystal will
bring the resonant frequency to the correct value, but won't allow the
oscillator to work.


It could bring it all the way to zero, on a bad day. The ESR of the
crystal is a factor of the overall loop gain; if the crystal is too lossy
then the oscillator may fail to start.

In general resistance in circuits like this bring the resonant frequency
down.


No. A checkbook helps in getting better components, but it's not a
passive component when used in this manner.

--
Tim Wescott
Control systems and communications consultinghttp://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes,http://www.wescottdesign.com/actfes/actfes.html

Hi Tim, thanks for the answer. The microcontroller I am using
(AT91SAM7S256) happens to have integrated load capacitances in the
leads of the oscillator, which ends up on a equivalent internal load
capacitance of 17-23pF which matches my Cload=20pF crystal.

I managed to find a nice crystal that provide me with the recommended
maximum ESR (for the frequency I selected for it).

Thanks...

 

[Franco]

Microchip has a good appnote worth reading.http://ww1.microchip.com/downloads/en/AppNotes/00826a.pdf

Cheers

Hi Martin, thanks for the doc. I think that is what I needed.

Thanks.

 

[Eeyore]

Hi everyone:

I was wondering if the external capacitors for matching Cload are
"always" neccesary when attaching a crystal to a microcontroller? If
the internal capacitance of the microcontroller oscillator is the same
as the specified in the datasheet of the crystal, then I don't see
reasons for the external capacitors (those two ones connected between
the the leads of the crystal and ground).

It would be very unusual for the stray capacitance to be that high.

I go by the microcontroller manufacturer's recommendations.

Graham

 

[Franco]

It would be very unusual for the stray capacitance to be that high.

I go by the microcontroller manufacturer's recommendations.

Graham

Ok, thanks for info.