Post Go back to editing

ADM2682 Receiver inputs A,B

Category: Hardware
Product Number: ADM2682E


I am testing a PCB which has this chip on it and I get weird measurements for the inputs A and B. The lines A,B,Y,Z are simply connected to a connector where a converter is connected and then goes to the PC. It is working fine (some boards have issues with the RS485 but I don't know yet if it is a firmware issue) and I don't know how with these measurements (1ms/div): 

Line A 

Line B

Have you ever seen something like that ?

The lines Y,Z seem normal (not a non-square shape). 


  • Hi BRSJoe,

       It appears that the other transceiver's controller is operating with a scheme to reduce operating power.

       With RS-485, there are 3 different voltage levels: A driven > B, B driven > A and "idle" with A>B.  Note that both A driven B and "idle" will map to a logic "1" output.

       Looking at your measurement, the leading and trailing portions would be when the RS-485 driver is "idle" and the VA & VB voltages are set by an external failsafe biasing network and the output is "1".  The RS-485 word starts with a leading "0" which is what is seen with the single step of the B driven > A level.  After the "0", there is a "1" which starts with a short period of A driven > B followed by the driver turning off and the voltages dropping to the idle state until the next "0" where the bus goes to the B driven > A.

       Because the RS-485 bus has a terminating resistor connected between A and B, the larger voltage difference of A driven > B over the idle state means a higher power dissipation.  Thus the driver is initially driven to the "1" state under full power to charge the bus parasitic capacitance and then turned off to save power and allow the fail-safe biasing network to maintain the idle "1" state. 

      In terms of if this is causing your communication problems, if the falling edge from A driven > B to the idle state is fast enough, it could cause the ADM2862 to incorrectly output a brief 0.  This can occur when the bus capacitance is very small, such as when using a short RS-485 cable (< 10m).  If this is the case, then adding a small amount of additional capacitance (~1.47nF) between A & B, will prevent this glitch.


  • Thank you so much for your quick reply, I'll check that.