• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 81 subscribers
  • Views 1484 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.

MSP4305419A and GSM Power and Interference Issues

Nithin James
Prodigy 160 points


Hi,

We are working on MSP430 and GSM based automotive application. In our current design we are facing following issue.

 - Crystal getting failed randomly during GSM activity with network (SIM registration, data transfer etc).

We are not able to find out the root cause for this issue. We are suspecting this might be due to Poor Power supply Isolation, EMI or EMC. Kindly suggest us how to go ahead with this issue.

Solutions we have tried:

1) We have introduced Snap Ferrites in the Power Cable from Vehicle Battery - No result

2) Ferrite beds between the GSM Ground and uC Power supply GND - No Result

3) Restart the uC when EXT OSC Failed - No Result

4) Changed the basic XTAL to AEQ-100 Spec XTAL - Little better results. Failure rate is reduced.

We doubt the following reasons for failing,

1) Crystal failure due to EMI/EMC

2) uC PC Corruption due to EMI/EMC

3) Failure due to Power supply surges due to sudden GSM bursts

Observations:

1) OSCFAULT Flag is set during SIM Registration

2) We dont have oscilloscope, we tried in PC based oscilloscope (20-20KHz) for XTAL (16 MHz) pins. We are able to see some burst noises during GSM Activity.

Kindly help us for the following:

1) Suggest some reference power supply designs with MSP430 and GSM for Automotive applications. (We are refining this with new design)

2) Suggest us how to isolate the GSM and Controller part.

Thanks in Advance,

Nithin James.

  • 0
    Guru 12050 points
    Tried 2 x 100uF tantalum caps to the modems power input?
    Select a lower power mode for 850-900band, as at full power it uses burst of 2A.
  • 0
    TI__Prodigy 100 points
    Hi Nithin

    Perhaps one of these helps:

    - Separation of GSM- and Controller-PowerSupply as best as possible, i.e. by implementing separate GND-Planes (if on same PCB) and blocking caps as near as possible to the controller Power pins. Allways use two caps (i.e. 1uF+100nF, depending on controller, please see Spec) on every Supply pin (every Analog and every digital Supply pin). The 100nF is the one that has to go nearest to Controller (No-Vias allowed!).

    - Implement Guard-Ring around the XTAL; normally used for 32kHz-XTAL, but perhaps usefull in your case: www.ti.com/.../getliterature.tsp;fileType=pdf&keyMatch=SLAA322&tisearch=Search-EN-Everything

    - Shield XTAL and perhaps also Controller (EMI-Shielding Cans).

    For further assistance the following information would be usefull, but of cource I understand disclosure issues:

    - PCB Layout, specially MSP430 and its Crystal and power supply connection.
    - Description of signals going around the XTAL
    - Clock-System Settings in Controller (XTAL-Settings)
    - Description of XTAL used and the used Load Caps
    - Scope shots of VCC and XTAL Signal (Clock switched out via GPIO) * I know that you don't have a scope available, but this information would be usefull...

    Hope this helps!

    Kind regards,
    Sebastian
  • 0
    Guru 23000 points

    Nithin James said:
    2) Ferrite beds between the GSM Ground and uC Power supply GND - No Result

    This makes it only worse, (high) voltage level between both grounds can occur, and so shifting the signal levels. Ferrite Bead’s needs to be used to block spikes coming from a device and must be placed very close to the source (device) as and in conjunction with the 100nF capacitor.

    Long wires/traces, and too long wires forming a coil, have the same effect.

    Ground wires/traces must use only one junction and forming a ‘Star’ connection. If your GSM module is grounded through your PCB (trace or plane) to the main power, than this can be a reason for the error.

  • 0 in reply to Sebastian Stumpf
    Prodigy 160 points

    Hi sebastian,

     As per your above reply we are willing to share the schematic and Gerber files for  verification to find out the root cause.  please share your official mail id so that we can share the details. My mail id is nithinjames@trinitymobility.com

  • 0 in reply to Nithin James
    TI__Prodigy 100 points
    Hi Nithin,

    I will do so.

    Kind regards,
    Sebastian

**Attention** This is a public forum