Post Go back to editing

LTC2872 Trying to confirm correct wiring for switching between RS-232 and RS-485 half duplex on one transceiver block

Category: Hardware
Product Number: LTC2872

I am excited to include the LTC2872 into a new design.  This is a late addition to a much larger system design and I am trying to confirm my understanding of the datasheet for my intended purpose.   After spending many hours with the datasheet, I contacted tech support but unfortunately stumped that very smart engineer and have been waiting many days for an answer.  This is critical path for releasing the bigger project board to proto production and here we are at the end of Friday still waiting for a confirmation.  Maybe you can help...

Let's just talk about one transceiver block and I will duplicate things on the second transceiver block.  (so instead of, for example, writing "TE885_1", we will just write "TE485")

OBJECTIVES: 

  1. Implement (two, one per transceiver block of the LTC2872) software switchable RS-232 and RS-485 half-duplex ports sharing common input and output lines.  "485/232bar" pin will select the active protocol mode.  (We will control what wiring is on those connectors.  Don't worry about that .. The configuration is set once at installation time and should remain the same for a long time unless user needs to change the sensors connected to the system and then we have to reconfigure the firmware anyway at that time  )
  2. Data to/from processor will use one RX line (pin) and one TX line (pin), respectively, for both RS-485 and RS-232
  3. Output from LTC2872 will share two pins to the real world... RS-485 (typically designated "A" and "B") and RS-232 (typically designated "Tx" and "Rx",  That is, the physical connector will have one pin used as either RS-485 "A" or as RS-232 "Rx", and one pin used as either RS-485 "B" or as RS-232 "Tx".
  4. RS-485 half duplex termination resistor will be selectable via TE485 pin on LTC2872.
  5. DXEN and RXENbar will be driven by processor when using RS-485.
  6. Half Duplex mode of RS-485 will be selected by H/Fbar.
  7. Fast Enable mode is not needed and will be tied to turn off for this implementation.

While it would seem that this is a typical set of objectives for using the LTC2872 and there are lots of "typical" examples given in the datasheet, this set of objectives is not illustrated in any of those examples, as far as I can see.

I believe the following color-keyed illustration and schematic *might* be the correct wiring to achieve these objectives but I am not ready to commit the board until I have someone familiar with the LTC2872 confirm this.  (And as mentioned above, my hours with the datasheet and time with the Analog support engineer so far have left us both uncertain).

One other note...  I want to be certain that we don't end up with "double" termination resistors with this wiring when TE485 is enabled.

Your time and assistance are greatly appreciated.



added "One other note..."
[edited by: PeteDD at 10:35 PM (GMT -5) on 27 Jan 2023]
Parents
  • Hi  

    I am @mentioning a couple of people here that may be able to help.      

    Please keep in mind, as mentioned in the Analog Devices EngineerZone Code of Conduct, ADI members strive to respond to your questions within three business days although sometimes a resolution to your question may take a bit longer.

  • Well no one every addressed this and after over a week, I also gave up on Analog's technical support.  I moved the design to a competitor's solution which in the end cost less and was more flexible.  Boards are in production.   Embarrassed for Analog.

  • Hi  ,

    I would like to apologize for the delay.  We are working to update our internal reporting tool for this product line and the responses for support via Engineer Zone have been delayed.  We are updating this now and future responses will be more prompt.  

    As IsoKeith mentioned, the configuration you propose may have an impact on the common mode performance due to the doubel termination.  In addition, I would recommend the following things:

    1. There may be an error in the VDD (pin 20) and VCC_2 (pin 21) connection.  The VDD pin is a regulated output at a typical voltage of 7V, so this pin should not be directly connected to +3.3V.  I think the intention was to have VCC_2 (pin 21) pulled up to +3.3V.

    2. Ensure that all VCC, VL, and VEE pins have individual decoupling caps placed close to the part.

Reply
  • Hi  ,

    I would like to apologize for the delay.  We are working to update our internal reporting tool for this product line and the responses for support via Engineer Zone have been delayed.  We are updating this now and future responses will be more prompt.  

    As IsoKeith mentioned, the configuration you propose may have an impact on the common mode performance due to the doubel termination.  In addition, I would recommend the following things:

    1. There may be an error in the VDD (pin 20) and VCC_2 (pin 21) connection.  The VDD pin is a regulated output at a typical voltage of 7V, so this pin should not be directly connected to +3.3V.  I think the intention was to have VCC_2 (pin 21) pulled up to +3.3V.

    2. Ensure that all VCC, VL, and VEE pins have individual decoupling caps placed close to the part.

Children
No Data