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.

Crystal Oscillator on the board and no bypass caps

Other Parts Discussed in Thread: MSP430G2252

Hi,

so I am trying to make a small pcb board with the MSP430G2252 DIP 20 package. I had a few questions in mind while I looked at the launchpad board.

I see that all the micros have programmable bypass caps in them for the 180 shift, then why does it have space for caps on the board and the ends of the caps don't even go to ground.

Second whats with the resistors in series with the outputs. I don't understand whats the point on even having these as outputs.

Please be descriptive.

Thank you 

Edit: One more thing, why are there so many unnecessary vias everywhere on the board. Its not connected to anything is it for heat conduction?  

Thanks 

  • Tashfique Abdullah said:
    I see that all the micros have programmable bypass caps in them for the 180 shift, then why does it have space for caps on the board

    So you have possibility to use external caps - if needed. Note that choice of internal capacitances is very limited - three values to be exact. What if your crystal needs 9pF? If you need just to play with LaunchPad then use internal XCAPs and you'll be fine.

    Tashfique Abdullah said:
    and the ends of the caps don't even go to ground.

    Look closer :) There's top layer ground plane below crystal and cap pads are connected to it.

    Tashfique Abdullah said:
    Second whats with the resistors in series with the outputs. I don't understand whats the point on even having these as outputs.

    If you don't use oscillator then those pins are at your disposal as GPIO pins. Note that if you do use oscillator, then better remove (unsolder) those 0-resistors

  • Ilmars said:
    Look closer :) There's top layer ground plane below crystal and cap pads are connected to it.

    Well I am looking at the PCB drawing on this upverter.com thingy and there are no trace going to the crystal casing or the caps. But the schematic is connected to ground. Something is wrong.

    Thanks for clearing all that out and that they are 0-ohm resistor just as jumpers.

    One more thing, does the msp have an internal oscillator, if it does, whats the freq and the point of including an external one. 

    Also, I am using this oscillator with 12.5pf , so I would need 23 pf capacitance, is that alright if I am off by 1 pf

  • Tashfique Abdullah said:
    One more thing, does the msp have an internal oscillator, if it does, whats the freq and the point of including an external one. 

    Yes, MSP430G2552 have on-chip oscillators. Why do you ask those question in forum instead of just reading them in product information - webpage and datasheet?

    Basic Clock Module Configurations

    • Internal Frequencies up to 16 MHz With Four 
      Calibrated Frequencies
    • Internal Very-Low-Power Low-Frequency (LF) Oscillator

    To know frequencies, please refer to datasheet

    Tashfique Abdullah said:
    Also, I am using this oscillator with 12.5pf , so I would need 23 pf capacitance, is that alright if I am off by 1 pf

    Maybe yes maybe no. Anyway you will get just frequency error, oscillator will work indeed. You shall take in account trace and i/o pin (parasitic) capacitances so maybe on some PCB using 12.5pF it's too much and 10pF is better choice. Anyway you shall check oscillator frequency using frequency counter and see resulting frequency error.

  • Tashfique Abdullah said:
    One more thing, does the msp have an internal oscillator, if it does, whats the freq and the point of including an external one. 

    The answer to your question is in the User's Guide and Device Datasheet, both available from the product page for the MSP430G2252 <- Click this

    Tashfique Abdullah said:
    Well I am looking at the PCB drawing on this upverter.com thingy

    If this is a TI dev board, then you should look at the drawings that you download from the TI website, not some third-party website where you have no reasonable expectation that information is accurate or up-to-date.

  • Tashfique Abdullah said:
    I see that all the micros have programmable bypass caps

    Actually, these are called 'load' capacitors/capacitance, not 'bypass'.

    Yes, almost all MSPs (not the FR5 family) do have them.However, the capacitance is not selectable on 1X family, and even on the other MSP families, it may not match the required load capacitance for your crystal.
    Also, the internal caps are not suited for HF operation. They are optimized for low power low frequency, but may fail due to HF effects above the LF operating range (up to 50kHz).

    So in case of HF crystals, you will definitely need external caps.

    Tashfique Abdullah said:
    Second whats with the resistors in series with the outputs. I don't understand whats the point on even having these as outputs.

    The resistors are 0Ohm resistors. Bridges. If you want to use the two XT pins for I/O, you'll need them to access the breakout connector. If you want to use the crystal, you have to remove them so that the additional traces and the breakout pins don't act as antenna and influence crystal oscillation.

    The LaunchPad offers both uses. The choice is yours.

    Tashfique Abdullah said:
    One more thing, why are there so many unnecessary vias everywhere on the board. Its not connected to anything is it for heat conduction?  

    They aren't unnecessary. They connect the shielding copper on the upper side with the GND plane on the bottom. This mostly eliminates crosstalk between the signal lines. Providing breakout connectors always calls or problems with EMI. Shielding each trace form the other reduces possible problems while doing experiments with a PCB that doesn't have an optimized layout for the specific experiment.

  • Thanks to all the answers now I am starting to get this.

    I have one more question,

    I have a device for detecting various colours and I can set a threshold in the device, its a I2c device, and it sends a interrupt signal through an interrupt pin.  

    The output is open drain, so I guess I am looking for a zero at the input as the interrupt flag. ... the input pin of the micro will provide the 3 volts right? Is it a cmos input? yeh I think I need a pull up resistor to vcc, coming to think of it or else how will it rise to vcc.

    Oh and what input pin can act as an interrupt flag? I am kinda confused looking at the data sheet. I am using a crystal and I2C so P1.6 , P1.7 and XIN and XOUT are all used up. Are there other pins to provide interrupt flag?

    Thank you for all the help.

     

  • Tashfique Abdullah said:
    the input pin of the micro will provide the 3 volts right?

    No. You need an external pullup resistor to the MSP430's Vcc voltage. You can look up the MSP430's input pin capacitance (and the I2C device's pin capacitance, add them together) and use that to size the pullup resistor for the edge rate that you want.

    Tashfique Abdullah said:
    Oh and what input pin can act as an interrupt flag? I am kinda confused looking at the data sheet. I am using a crystal and I2C so P1.6 , P1.7 and XIN and XOUT are all used up. Are there other pins to provide interrupt flag?

    Any of the Port 1 and Port 2 I/O pins can be programmed to provide an interrupt on external signal edge. Read the section of the User's Guide (8.2.7) and Interrupt Vectors section of datasheet.

**Attention** This is a public forum