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.

TS3A24159: RS485 bus switching issue..

Part Number: TS3A24159
Other Parts Discussed in Thread: TMUX136, TS3USB3031, TMUX1072

Hi,

In my project i am using TS3A24159 SPDT switch for RS485 bus switching application. My design RS485 slaves are connecting with daisy chain method. Each slaves are switching the bus to next slave. Here i am facing an issue of data loss when i connecting 3 or more slaves. Can you please suggest a solution for this.

Thank you..

  • Monu,

    The TS3A24159 device is only a 2:1 switch so I'm curious how you are connecting 3 or more slaves. Do you have a diagram of what you are trying to do?

    Also what is the signal voltage range you are using for RS485?

    Do you have a scope shots of the input of the switch to the output of the switch when you are seeing data correctly and when you are seeing incorrect data?

    Another thing you can try is removing the mux and shorting the signal path to see if your TX and RX are working properly.

    I wonder if you have too much capacitance on the bus with all the devices on a single bus and the TS3A24159 is a low Ron device so it a large has on-state capacitance of over 200pF. Maybe a lower capacitance switch like the TMUX136 could solve the issue but we won't know until we see what is going on with the waveforms passing through the switch.


    Thank you,
    Adam
  • Adam,

    We already remove the device and test the communication is working fine.

    The voltage level of the RS485 is 0-2V

    All the slave have the TS3A24159 device, and each slave enable bus for communicating to the next slave. The bus series resistance will increase through each device.
  • Deepak,

    Do you have a scope shots of the input of the switch to the output of the switch when you are seeing data correctly and when you are seeing incorrect data?   This could help see if there is too much capacitance on the bus if we see slow rising edges compared to the signal when the switch is removed and the trace shorted.

    I don't understand your signal path connections you describe do you have a diagram?

    What RS485 master and slaves are you using?   

    In this configuration using a device with large capacitance 200pF per switch one of the slaves signal path could see as much as 400pF.  There could be other solutions available using devices with low capacitance like the TMUX136 or 3:1 switch like the TS3USB3031. 

  • Adam,

    Here i am sharing the connection between the master and slave with bus switch.

    RS485 transceiver  ISL83078E which is Renisas chip.

  • Deepak,

    The TS3A24159 is not an RS485 slave device so I'm not sure what you are describing.  Are you trying to show that you are daisy chaining RS485 slaves?  

    Are you able to provide a scope capture or use the TMUX1072 which is a lower capacitance device?  I'd like to rule out if the high capacitance of the TS3A24159 is the root cause of your issue.  

    Thank you,

    Adam

  • Adam.

    Sorry for the miss understanding, actually our plan is to connect 100 nodes on the RS485 bus. now we build only 4 boards, when i am connecting this 4 board in the bus it happening the data loss. Now we don't have any wave forms because of the lack of oscilloscope. if i am using   this part TMUX1072 instead of this TS3A24159. Is this resolve the issue?, here i am sharing the connection diagram. This configuration is for to achieve dynamic addressing of the slaves. Initially the switch is OFF so the master and 1st slave only have path to communicate, after the addressing of 1st slave activate the switch ON and then the master can communicate with the next slave. This mechanism is repeatedly doing till the last slave has been addressed.  So can you please suggest any solution as soon as possible.

      

  • Deepak,

    Without seeing the waveforms it is hard to say how you are observing data loss and what might be the root cause.  Without knowing the root cause I can't say if the TMUX1072 device will solve your issue it is simply a guess that you have a lot of bus capacitance in your signal chain and the transceiver cannot drive it at the speed you are attempting.

    I will notify our RS485 team to see if they have any comment.

    Thank you,

    Adam

  • I agree with Adam that observing the waveforms would be the best way to debug this. If that's not possible, though, then another thing you could try is reducing the baud rate on the RS-485 bus. (By the way, at what baud rate are you currently testing?) If communication works at slower rates, it may point to high total bus capacitance as a likely root cause.

    Am I correct in understanding that you confirmed the communication works between all four boards when the switch is removed/bypassed? (I wasn't sure if you meant you removed just the switch or if you removed the whole node.)

    If you can share a schematic of the interface implementation for both master and slave boards it might help as well.

    Max
  • Hello Max,

    Here i am attached the updated schematics with TMUX1072 switch.  And also we tested with 4800 baud rate, now it improve the communication.

    smart rod_slot_sch_reva_29mar20.pdf

  • Hi Deepak,

    This schematic shows a 120-Ohm resistance (R12) between the A and B lines of the RS-485 bus. Is this resistance populated at every node? If so, then the total effective bus resistance may be dropping too low and causing the differential voltage amplitude to drop. (RS-485 transceivers are typically design to drive buses with two such termination circuits, giving an equivalent load resistance of 60 Ohms. If you had this termination at four nodes, though, the resistance would drop to just 30 Ohms.)

    Max
  • Hello Max,

    The 120-Ohm resistance is only mounted on Master and the last connected slave. in between it will be demount .

    Thank you.

  • Deepak,

    OK, that should not be an issue then. Do you have any way of observing the signal waveforms so that we can better understand what might be going wrong?

    Max
  • Yes right now we didnt have the wave form. We will update it soon.