Post Go back to editing

MAX14850 I/OA1, I/OA2 with UART

Thread Summary

The user plans to use a 4.7k pull-up resistor on both ends of UART RX/TX lines for RS-232 communication through the I/OA port of the MAX14850. The support engineer suggests potential issues with the 22 ohm series resistors affecting the output low voltage and the rising edge speed, and recommends considering the ADuM363N as an alternative. The engineer advises checking the net capacitance and data rate to ensure the rising edge is fast enough.
AI Generated Content
Category: Hardware
Product Number: MAX14850

I plan to configure the circuit as follow : 

connecting a 4.7K pull-up resistor to both ends of the UART RX/TX lines for RS-232 communication through I/OA port.

Will there be any issues?

I'm concerned because when I previously connected through the I/OA port using SPI, there were a lot

of issues. All the other IN/OUT ports of MAX14850 are already being used elsewhere.

  • Hi AndersonKim,

       Without seeing the rest of the design to see what and how the MAX14850 is connected for these signals I can't definitively say if there will be issues or not, but what you have shown seems like a plausible starting point.  As an alternative option, I'm wondering if you have considered the newer ADuM363N.  This newer family of digital isolators will give you 6 standard digital isolation channels without having the complexities of the MAX14850's two bi-directional channels and associated pull-up resistors.

       Assuming you want to stay with the MAX14850, here are my comments/potential concerns you may need to address with respect to the bi-directional IO channels.

        The 22 ohm series resistors are a source of potential concern for me.  I believe that usually they're added to digital circuits as a form of series damping/rise time control to help minimize the issues associated with fast edge rates.  Typically if fast edge rates aren't actually a problem, the additional series resistance doesn't cause problems, so they a low-risk addition for the common fully driven logic outputs. 

       In this instance, the IOAx and IOBx ports have an open-drain output and use the net's effective parasitic capacitance in conjunction with the pull-up resistor to generate the rising edge.  This is typically much slower than the rising edges achieved by most logic device outputs which is why the series resistors are likely not needed here.   These resistors do, however, add some series resistance which contributes to a voltage divider effect that, depending on their position in the circuit relative to the net's driver and receiver and the pull-up resistor(s), can effectively increase the output low voltage observed by the net's receiver device.  In some cases, this incremental increase is sufficient to cause the receiver to not recognize the output low voltage being driven on the net.  Bidirectional isolators usually have a higher output low voltage on one side as part of their mechanism to prevent forming a latch and this higher voltage VOL can worsen this issue.

    This scenario is much more common in I2C buses where some parts have a low VIL level that can be very close to another device's high VOL level.  I don't know that this will be an issue for your circuit, but without reviewing the details of the other devices, I can't say.

    The other potential point of concern is if the rising edge generated by the 4.7k pull-up will be fast enough for your bit rate.  It probably is, but I don't know the total effective net capacitance or your data rate which prevents checking if this is an issue or not.

    Eric 

  • I'll set 115200 bps for 232 communication. Thank you for your response. I will keep it in mind.

Before You Switch


Switching languages will make ADI Explorer unavailable. Resume your session by switching back to English and reopening ADI Explorer.