• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 79 subscribers
  • Views 1093 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

External XT2 crystal oscillator, will this work?

Emil Svensson
Prodigy 40 points

Hi so I'm doing some work with the msp430F5438 experimenter board and for good timing and fast transmission threw the USART usb I need a crystal oscillator.

The timer will need to run at  3.2KHZ intervals with good precision, to clarify the precision the application I'm developing will trigger a ADC12 transfer for a period of MAX 10min most likely less, so some error is acceptable.

So from what I've gathered I need 1 crystal and 2 capacitors for it to work. So I've picked out the following parts:

http://search.digikey.com/us/en/products/9B-20.000MEEJ-B/887-1248-ND/2207668

http://search.digikey.com/us/en/products/C1608C0G1H070D/445-5041-1-ND/2093656

So the values I've been looking are the following:

Crystal

Frequency tolerance/stability  +-10ppm        

Shunt Capacitance                    7pF MAX

Load Capacitance                      18pF

Frequency                                    20MHZ

Capacitors

Dimensions                               0.063" L x 0.031" W (1.60mm x 0.80mm)              Should be right according to the parts list

Capacitance                               7pF

So my questions is will this work if soldered these onto the experimenter board?

Should I reconsider any of the parts?

If you have suggestions I'm very glad to hear them, the specific frequency is not that important. I was planning to directly use it to source SMCLK and MCLK and since the CPU can handle up to 25MHZ I would like it to operate in +20MHZ. Power consumption is not an issue.

  • 0
    Guru 227245 points

    Emil Svensson said:
    Shunt Capacitance                    7pF MAX
    Load Capacitance                      18pF
    [...]Capacitance                               7pF

    no, the capacitor is wrong.

    The Shunt capacitance is th einternal capacitance of the crystal at resonance frequency. The crystal electrically acts as a capacitor, just that it stores the electrical energy by mechanical deformation rather than by building up an electron cache. On teh mechanical resonance frequency, the capacitance is biggest, so the electrical oscilaltion maximizes.

    However, to make this work, a 'load' is required. This load is 18pF parallel to the crystal. But the way the Pierce oscillator in the MSP works, this load is split into two capacitors which are in series to each other, with GND connected to their middle point. Series capacitors are calculated like parallel resistors. So you need two capacitors of 36pF each to form a 18pF load. However, the MSP pins by themselves have 2pF each on parasitic capacitance. And if you have a socket...
    Well, we use 36pF capacitors on our devices too. Sometimes even 39pF, if for some reason the crystal won't start (which sometimes happens for two years now, even on devices on which we neither changed layout nor crystal type, but never happened before)

  • 0 in reply to Jens-Michael Gross
    Prodigy 40 points

    Thanks Jens-Michael for helping me once again.

    I spoke to a large electronic components distributor (whom shall rename nameless) where I spoke to a "technician" who claimed that the capacitors where suppose to match the shunt capacitance for it to work as I intended, I had explained what I was working with and so on. Thankfully they did not have any capacitors available for the crystal they had so I turned here before buying my parts online which was lucky.

    So I just wanted to say thanks for the quick reply.

  • 0 in reply to Emil Svensson
    Guru 227245 points

    Emil Svensson said:
    I spoke to a "technician" who claimed that the capacitors where suppose to match the shunt capacitance for it to work as I intended

    Well, he's right - but not for a Pierce oscillator. There are other oscillator circuits, all with their own drawbacks and advantages, and usually more complex, where indeed the external capacitance has to match the shunt capacitance.
    However, not when operating with the MSP oscillator circuit.

    Well, there is a chance that you got an overtone crystal. Those can have problems to oscillate, or may swing on a lower frequency (usually 1/3 of nominal), with the simple pierce oscillator. I actually don't know how to operate them safely, I just know that they exist and are sold with increasing likelyhood, the higher the nominal frequency is (never seen on on 10Mhz or below, but above 25MHz, many offered crystals are overtones).
    You should check this.

**Attention** This is a public forum