Post Go back to editing

RS485 output does not swing from 0V to Vcc wrt GND

Category: Hardware
Product Number: MAX13487E

Hi, I have an Third party device with RS485 interface implemented using the above IC. We normally expect Va and Vb to swing fully from 0V to Vcc. But in this case Va swings from Vcc/2 to 0V and Vb swings from Vcc to Vcc/2. In other words like following diagram.

The top most is the expected RS485 behavior and the middle is what I see from the device. Bottom is Vdiff (Va - Vb).

As we can see, due to this behavior the diff voltage is 0V for logic '1', which is in undefined range for a RS485 receiver. 

The particular implementation has termination of 120R and fail safe resistors of 3k3. nRE and nSHDN are tied together and pulled high. I have not connected anything else on the bus. 

Can someone explain this? Or have I make some mistake in my understanding?

Thanks,

kaushalya

Edit Notes

explanation about the diagram added
[edited by: kaushalyas at 6:02 AM (GMT -4) on 8 Apr 2025]
  • Hi Kaushalya,

    The MAX13487 is an auto direct RS485-compatible transceiver. Unlike the typical RS485 transceiver the MAX13487 features internal state machine using DI and A-B signal to manage the local receive and transmit enable pins. The pull-up and pulldown resistors on the A and B lines are required for proper operation. We have an Ev kit for the is part, and the pull-up and pull-down resistor value on the EV kit is 220ohm. The 220ohm biasing resistors should provide a balanced VOD without any load. Regarding the output voltage swing on A and B, please see the figure (figure 4 from the datasheet) below for the transmitter output levels, 0V to output normal high/Vcc to output normal low.

    For auto direct part, it is common to see VOD waveform have an offset below 0V since the receiver positive threshold can be shifted below 0V to make the output stay high even the differential voltage on the bus equals to zero.

    Best,

    Jane

  • Hi YaoJ,

    Many thanks for your reply. the A and B pull up/down resister in this particular design is 3k3, which is much higher than 220 ohms.

    1. Do you think that could affect the transmission?

    2. How can you derive the max values for these pull up/down resistors?

    3. So you say that  Vod can be 0V while in transmission? i.e. Vod changes between 0V to a negative voltage while data transmit. 

    4. I dont have the EV kit with me. Would you be able to capture a scope trace of a data transmission via the EV?

    5. Do you have a SPICE model for this IC? 

    Cheers,

    Kaushalya

  • Hi Kaushalya,

    The VOD waveform depends the resistive and capacitive load on the bus. The value of these resistors set the idle voltage. Please request a MAX13487 EV kit and evaluate it with your application. See below screenshot from the MAX13487E EV kit, using 220ohm biasing resistors without any load. When a 120ohm termination is added, the VOD will shift down. I fed a signal at TXD and monitor A and B on the EV kit (C1=TXD at 100kHz, C2 =A, C3=B, and M=VOD). Unfortunately, we do not have a SPICE model for this part yet. Please let us know if you have any question regarding the EV kit or would like to choose a different part for your application.

    Best,

    Jane

  • Hi Jane,

    Thanks for that. 

    Yeah this is what I would expect the RS485 transceiver would work, but in my case Vod only goes from 0V to -3V, which is not going to work for my other devices. Unfortunately this device is not our design as it is a third party device we buy off-the-shelf. 

       How can you derive the max values for these pull up/down resistors?

    Thanks,

    Kaushalya

  • Hi,

    I guess the pull up/down resistors only have to keep the Vod > +200mV when idle. Based on 120R termination and 5V supply, I see the pull/up down resisters should be less than 1.4k. This is of course not considering any other fail safe resisters on the bus. Is this correct?

    What happens if pull up/down resisters are more than 1.4k?

    Cheers,

    Kaushalya

  • Hi,

    I have replaced the 3k3 with 200R resistors and situation is much better. This is without any other devices on the RS485 bus.

    Following is the scope capture.

    Yellow : Va

    Green : Vb

    Blue : Va - Vb or Vod

    As you can see from Vod trace, it verris  between +1V and -2.7V. This is inline with RS485 standard. But I like to know why is it asymmetric?  It seems like there is a common mode voltage of (1-2.7)/2, or -850mV. What is the reason for this?

    Thanks,

    Kaushalya

  • Hi  Any update on this?

    Thanks

    Kaushalya

  •  Hi Kaushalya,

    Please see the link for an apps note for the MAX13487. The swing of the driver output designs for isolation with optocouplers and applications with auto direction signal at DI. The AutoDirect state machine monitors the voltage level on the bus (A-B) to determine the state of the driver. TI also has a good apps note explaining Auto-Direction applications.

    https://www.analog.com/media/en/reference-design-documentation/design-notes/ds96-simplify-your-industrial-interface-by-eliminating-rs-485-control-lines.pdf

    https://www.ti.com/lit/an/slla574/slla574.pdf?ts=1744227685611

    Best,

    Jane

  • Hi Jane,

    Thanks. I didnt quite get what you are saying "The swing of the driver output designs for isolation with optocouplers and applications with auto direction". The IC can be used in non-isolated environments as well. Only thing with Auto direction control is you dont need an additional isolator in case you implement isolated design. 

    As I mentioned in my previous comments, the line has to be kept above +200mV when idling for the Auto direction control to work. Please let me know if I am correct.

    Cheers,

    Kaushalya

  • Hi,

    Ok I think I figured it out. This chip floats the bus when transmitting logic '1'. So fail safe resistors are mandatory to keep the bus above +200mV when floating. That is how logic '1' is transmitted.

    Is my understanding correct?

    Cheers,

    Kaushalya