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.

ISO3086T: RS485 interface only working in one direction

Part Number: ISO3086T
Other Parts Discussed in Thread: SN65HVD72, , ISO3082

Tool/software:

Hi,

I use SN65HVD72, SN74LVC2G14DBV & ISO3086TDW design a isolated RS485 interface between my IPC & device(laptop).

But unfortunately my design only working in one direction, IPC to device.

I test by using Tera term and found that there is no communication from device to IPC.

RS485 interface issue.xlsx

  • Hello Lee,

    Thank you for reaching out. At first glance, I think this is an issue with how RE/ and DE are tied together. These pins are commonly tied together in many applications on order ensure that the ISO3086T is either transmitting or receiving. In order to enable communication from device to PC please make sure that DE is enabled using a high or "1" (RE/ will be enabled on a low or "0").

    Let me know if this fixes your issue. 

    Best,
    Andrew

  • Hi Andrew,

    Thank you very much for the reply.

    IPC communication to device OK, 

    R =1.9V, RE =DE =3V , D= 400mV

       

    when communicate from device to IPC, A & B receive signal from device.

    But there is no any respond of pin R, RE, DE & D.

    R=1.9V, RE=DE=3V, D=400mV

    Look like my circuit design always stay at IPC to device stage.

    Thank you.

    Best regards,

    Ku

  • Hi Andrew,

    basically my circuit design was follow Dual Isolated Half-Duplex RS-485 Repeater as reference.

    the only different was both ISO3082DW ic i change to SN65HVD72 and ISO3086T ic.

    Best regards,

    KU

    sn65hvd72.pdfISO3086TDW.pdftidu590.pdf

  • Hello KU, 

    What is the data rate of your application? The app note that you have mentioned uses the ISO3082 which means it is intended for much lower data rates. 

  • Hi Andrew,

    according to ISO3082 reference design summary features, these design data rate up was to 100kbps .

    The data rate i set in tera term was 9600.

    ISO3082 design feature:

    So in order to change the ISO3082 ic to ISO3086TDW ic for my design.

    i need to reconsider all the value of resistor and capacitor of my circuit to cope with ic ISO3086TDW?

  • Hello Lee, 

    The components should be the same since the data rate is not significantly changed from the reference design. Please allow me another day to review in detail. 

    Best,
    Andrew

  • Hi Lee, 

    Thank you for your patience and sorry for the delay. I understand you are using Dual Isolated Half-Duplex RS-485 Repeater as a basis for the design and ISO3086TDW as an upgrade to integrate the transformer driver. This is okay as long as the data rate matches the original TI design. 

    1. First, I would remove the 1kohm resistors R219 and R222 since these are not needed and only on the 'A' line. 
    2. Is see the ISO3086T is configured as a half-duplex since the Y/Z pins are to A/B pins to create one bidirectional bus for both transmit and receive. 
      1. ISO1452: [FAQ] How to configure a RS485 Full-duplex transceiver as Half-duplex?
      2. I think this may be the primary problem. J22 seems to have the A and B connections flipped (J22's B pin is connected to ISO3086T's A pin). Since the Z and Y connections are correctly connected, this would cause a communication error.  
      3. If the J22's A and B pin are disconnected this should allow you to confirm the applications receive function using the Y/Z pair which is connected correctly). 
    3. Otherwise, the I do see that the TI design is not matched exactly (for example, some passives are missing or the wrong value). This may compromise the timing of the discrete repeater. However please first confirm point #2 first. 

    Best,
    Andrew

  • Hi Andrew,

    I already removed 1kohm resistor R219 & R222 per your suggestion.

    Also modified my connection of J22 pin A connect to ISO3086T's A pin, and J22 pin B connect to ISO3086T's B pin.

    unfortunately the result still same , communication signal send from device (J22) to IPC (J23) still NG.

    But communication signal send from IPC (J23) to device (J22) is OK.

    I had capture the waveform for each IC pin during OK and NG condition for your reference.

    From the output waveform of J23 pin A & B, signal are distorted .

               

    Please refer to attached excel file.RS485 connection issue 1.xlsx 

    Best regard,

    Ku

  • Hello Ku,

    Thank you for the updated schematic. The design is still missing some of the critical components that set the timing for the auto direction. The output of 2A and 2Y should match the signal on the R pins with some slight delay which is used to switch the direction (using RE/ and DE) for each bit (as shown below).  

    1. Two 10k pull-up resistors are missing
    2. The 300-ohm resistors should be connected between the inverter inputs and the R pins. Then a capacitor to GND. 
    3. The 5.1k resistors are 50-ohms which would change the direction timing. 
    4. Section 2.2 Data Flow Control of the reference design goes into a lot of detail about the importance of the resistors.
      • After review, I realized that R219, R228 and R222 may need to be added back to the design later using the values in the design below. 
      • However, this can wait until we have corrected function of the auto detection circuits. 

     

    I have marked them on the schematic. 

     Best,
    Andrew

  • Hi Andrew,

    according to your No.2 comments :  300-ohm resistors should be connected between the inverter inputs and the R pins. Then a capacitor to GND. 

    It was different from reference schematic below. As can seen, there was no dot between 300-ohm  the inverter inputs and the R pins. 

      

    I will modified my design exactly same as the reference design by 

    -add 10k pull up to both the R pins.

    -change the 49.9-ohm to 5.1k-ohm.

    Will let you know the above result in my next reply.

    Thank you.

    Best regards,

    Ku

      

  • Hi Andrew,

    I already modified my schematic accordingly same as reference design,

    -add 10k pull up to both IC U10 R pin & IC U9 R pin.

    -change both R223 and R224 to 5.1k ohm.

    -change R225 & R226 to 300 ohm

    -insert R219 & R222 with 560 ohm

    unfortunately the result of RS485 communication signal from device (J22) to IPC (J23) still failed.

    According to waveform i took , the signal of U10 pin R, RE , DE and D look OK.

    But U10 output pin A & B waveform become distorted & suppose OK condition A& B signal are inverted,

    but now was not inverted.

      

  • Hello Lee, 

    The current state of your application is not particularly clear. As I understand the problem, Communication from J23 to J22 is play. However, communication from J22 to J23 has problems, correct? 

    I have reviewed the excel file and the communication from J22 to J23 looks acceptable until the output of A and B itself which looks to have some ground coupling issue. Looking at the schematic again I see that C314 is in series with the U10 ground pin. This would essentially float the ground connection and result in the device being unpowered.

    This was probably meant to be a decoupling capacitor (C314 connected across VCC to ground). U10 pin 5 must have a direct connection to ground. Is this fixed? 

    Best,
    Andrew

  • Hi Andrew,

     As I understand the problem, Communication from J23 to J22 is play. However, communication from J22 to J23 has problems, correct? 

    -Yes, only J22 to J23 communication had problems.

    U10 pin 5 must have a direct connection to ground. Is this fixed? 

    -When U10 pin 5 direct connect to ground, condition became more worst , communication from J23 to J22 also became NG.

    U10 become not working at all, when signal send from J23 to J22 / from J22 to J23.

    Input signal to U10 pin A & B but there is totally no output pulse from U10 pin R, (pin R =5V, RE=DE=0V , D= 5V).

    When signal send from J22 to J23, pin R=5V, RE=DE= 0 pulse, D= 5V pulse all look still OK at this stage. But there is no any output pulse being detected at U10 pin A & B.

    Regards,

    KuWaveform when short C314 to GND.xlsx

  • Hi Andrew,

    Since i almost try all the possible method but still not working for my communication from J22 to J23.

    and it look like mainly signal lost or distorted after go through U10 (SN65HVD72).

    Do you think is it i have use the wrong U10 ic?

    Thank you.

    Best regards,

    Ku

  • Hi Lee, 

    When U10 pin 5 direct connect to ground, condition became more worst , communication from J23 to J22 also became NG.

    U10 become not working at all, when signal send from J23 to J22 / from J22 to J23.

    I am thinking U10 may be damaged since a lot of testing has happened and signal was applied without VCC. This normally is okay since the device is protected, however, it is still outside of absolute maximum conditions. 

    • Please make sure U10 measures 5V from pin 8 to pin 5.
    • Please also solder a new unit to U10 and try again. 
    • C314 is also still needed as a decoupling capacitor between 5V from pin 8 to pin 5
    • It should be possible to try powering the logic side of the application using VCC = 3.3V since ISO3086T can operate on this. Just make sure VCC2 is still 5V

    It would be good to select a new device for U10, SN65HVD72. SN65HVD72 is a 3.3V device and VCC of 5V is technically outside of the recommended range. However, nothing about the datasheet shows me this device would immediately damage in the application. However, a RS-485 transceiver with VCC = 5V is preferred for this application.

    Also, since C314 was previously in series with Pin 5, this would effectively stop the device from being powered. I suspect previously the device was being powered through the ESD structure, which is not ideal, and maybe connecting the SN65HVD72 properly provided a VCC greater than 5V and damaged the device. 

    Best,
    Andrew

  • Hi Andrew,

    Finally my RS-485 communication working fine in both direction after

    -short C314 direct to ground (U10 ground pin 5 direct connect to ground)

    -SN65HVD72 , U10 Vcc change to 3.3V.

    Actually i try before change the U10 ic Vcc to 3.3V but communication from J22 to J23 still NG because U10 ground pin 5 was not direct connect to ground 

    at that moment , C314 was still inserted.

    Thank you very much for your great support to solve my design issue.

    Best regards,

    Ku

  • Hi Ku,

    Thank you for being patient and updating me on what worked for the final solution. I will close the thread now.

    Please feel free to ask a new question for your next design!

    Best,
    Andrew