Post Go back to editing

MAX3232 interface noise protection

Category: Hardware
Product Number: max3232ese, max3232e

Hi, ADI expert

The figure below shows the serial port design of my MAX3232, The board is connected to the PC through the serial port cable. During the test, I found that there were noise glitches in the phase noise curve of  other chip on the board, but after removing cable(at this time, it could not communicate with the PC, the board card could still work normally), there were no noise glitches in the phase noise cureve of the chip.

I think it's the noise that goes from cable to the board, so how do I design the protection circuitry on the MAX3232 to prevent noise from affecting the board's performance?

    

The picture on the right is trace of rs232_tx signal, PC -> cable-> CON1The picture on the left is trace of rs232_tx signal, no cable connecting to CON1.

    

Parents
  • Hi licuxu,

    The RS232 interface is not good at noise immunity, and the noise can be picked up from the surrounding environment through the cable. Shorten the cable length and using shielded cable improve noise performance. The longer the cable, the lower the data rate. The data sheet lists the guaranteed 120kbps data rate with CL = 1000pF.

    I am not sure what noise glitch in the noise phase curve you refer to. From your two waveforms, the ripple shows the charge pump changes switching frequency while the cable is connected to the board. The part is loaded with the long cable. It is possible you see different switching frequencies on the phase noise curve.

    Adding a cap closes to R_IN pins to ground could help to filter out high frequency noise, but it would also increase the capacitance load on the line.

    Best,

    Jane

Reply
  • Hi licuxu,

    The RS232 interface is not good at noise immunity, and the noise can be picked up from the surrounding environment through the cable. Shorten the cable length and using shielded cable improve noise performance. The longer the cable, the lower the data rate. The data sheet lists the guaranteed 120kbps data rate with CL = 1000pF.

    I am not sure what noise glitch in the noise phase curve you refer to. From your two waveforms, the ripple shows the charge pump changes switching frequency while the cable is connected to the board. The part is loaded with the long cable. It is possible you see different switching frequencies on the phase noise curve.

    Adding a cap closes to R_IN pins to ground could help to filter out high frequency noise, but it would also increase the capacitance load on the line.

    Best,

    Jane

Children
No Data