Author Topic: Crystal with "load capacitance" or not?  (Read 7047 times)

0 Members and 1 Guest are viewing this topic.

Offline onemilimeterTopic starter

  • Frequent Contributor
  • **
  • Posts: 314
Crystal with "load capacitance" or not?
« on: March 27, 2011, 02:56:27 pm »
I wish to purchase a crystal for PIC18F4431. I've two options as follows:

[ a ] XT9M-NL-A-NA-10M

[ b ] XT9M-18-A-NA-10M

These devices are manufactured by Vishay and the datasheet can be downloaded here: http://www.vishay.com/doc?35012

The designation of the part# is shown in attached image.

I'm confused with the "load capacitance". Should I purchase Part [ a ] or Part [ b ]?

If I use Part [ a ], does the oscillation circuit need C1 and C2 as shown in Figure 2-1?

Thanks.
 

Offline monpjc

  • Contributor
  • Posts: 32
    • monpjc
Re: Crystal with "load capacitance" or not?
« Reply #1 on: March 27, 2011, 04:13:19 pm »
Hi,

For reference to others: The PIC data sheets recommends a value of between 27pf and 22pf.

I'd go with the 18pF as specified by Vishay - this is how I've loaded crystals on PICs in the past and not had problems. Use the circuit as per PIC data sheet, that is two Cap's but no resistor, I've never needed it.

Would always check your osc on a scope and enable osc start up timer in the PIC (delays reset line while it becomes stable).

Also connect nice and close to chip with a good ground, not just a bit of wire or thin track, something solid.
 

Offline Thermal Runaway

  • Contributor
  • Posts: 48
Re: Crystal with "load capacitance" or not?
« Reply #2 on: March 27, 2011, 05:12:29 pm »
Use the circuit as per PIC data sheet, that is two Cap's but no resistor, I've never needed it.


@op The external feedback resistor monpjc talks about is used to aid oscillator startup.  It plays no further part once the crystal is up and running.  I normally do include it in my designs, but to be honest I have to agree with monpjc that I've never actually come across a situation where the crystal would not oscillate when the resistor is missing.  So, from a practical point of view, I can't disagree with his advice to omit it.

Here's a very good application note about the use of crystals by Microchip:

http://ww1.microchip.com/downloads/en/AppNotes/00849a.pdf
--------------------
Electronics Engineer, Land Rover enthusiast, Amiga Computer fan and general GEEK
--------------------
 

Offline onemilimeterTopic starter

  • Frequent Contributor
  • **
  • Posts: 314
Re: Crystal with "load capacitance" or not?
« Reply #3 on: March 27, 2011, 05:43:50 pm »
According to the datasheet, the load capacitance of part [ a ] XT9M-NL-A-NA-10M is not specified. Does it mean C1 and C2 are not required? Kindly advise. Thanks.
 

alm

  • Guest
Re: Crystal with "load capacitance" or not?
« Reply #4 on: March 27, 2011, 07:19:55 pm »
Series resonant crystals don't need load capacitance, but I don't think they'll work with the PIC18 crystal circuit, you would probably need to build an external oscillator circuit (I think the PIC16 datasheet contained an example) to use it. This webpage explains the difference and links to various other pages. Microchip AN826 explains it in more detail.
 

Offline onemilimeterTopic starter

  • Frequent Contributor
  • **
  • Posts: 314
Re: Crystal with "load capacitance" or not?
« Reply #5 on: March 27, 2011, 08:10:03 pm »
Thanks alm. The link is excellent.
 

Offline maggi9donald

  • Newbie
  • Posts: 2
Re: Crystal with "load capacitance" or not?
« Reply #6 on: April 05, 2011, 07:13:12 am »
According to the datasheet, the load capacitance of part [ a ] XT9M-NL-A-NA-10M is not specified. Does it mean C1 and C2 are not required? Kindly advise. Thanks.

C1 and c2 are required.

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13968
  • Country: gb
    • Mike's Electric Stuff
Re: Crystal with "load capacitance" or not?
« Reply #7 on: April 05, 2011, 08:17:12 am »
caps are always needed, although some ceramic resonators have them built in, crystals don't.

A useful tip I picked up at a Microhip seminar many years ago is that a 'happy' xtal osc has similar amplitude waveforms on each pin - note you need to use a low-capacitance probe like a FET or x100 - you can improvise with a 10m resistor on the top of a x10 probe as you are only comparing values so absolute accuracy isn't needed.
Bear in mind that there will be some capacitance from the PCB layout, and that load capacitance specced in a xtal datasheet is the series combination of the two caps, i.e. the external caps should be about twice the xtal spec, less stray capacitance.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline AdShea

  • Contributor
  • Posts: 35
Re: Crystal with "load capacitance" or not?
« Reply #8 on: April 11, 2011, 04:07:22 am »
Only time I haven't used load capacitance on a crystal is when I'm using a crystal with a tiny rated load capacitance (8pF) and the MCU oscillator pins already provide most of it (7.5pF).  In that case the board is enough to provide the rest.
 

Offline Leo Bodnar

  • Frequent Contributor
  • **
  • Posts: 811
  • Country: gb
Re: Crystal with "load capacitance" or not?
« Reply #9 on: April 21, 2011, 05:14:17 pm »
I have never used any xtal load caps on any of my designs having used anything from 4MHz to 16MHz on PIC16, PIC18, PIC24, dsPIC33.  I know Muntz wouldn't either.

I am not a big fish but 20,000 unloaded xtals are out there all oscillating happily - each tested by me personally.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf