1
\$\begingroup\$

I was testing the RS485 Circuit, below is my circuit and I connect J5 to my PC using another RS485 to USB converter. In PC through Modscan software I was monitoring the polling and response, a twisted wire used, length of wire is 3 feet only. enter image description here I observe the polling missing some time, generally out of 100 polls, I missed three response. my PC is master and device is Slave. So I probe the waveforms and observe while in receiving mode I saw a lot of noise on RS485 waveforms. and when transmitting mode such noise is not there. see the attached waveform, does any have any idea, this is how an RS485 behave. any suggestion, advise.

I have two channel scope only. enter image description here

enter image description here

layout ( 4 Layer PCB) enter image description here

enter image description here

\$\endgroup\$
14
  • \$\begingroup\$ How are you testing this circuit? Did you layout a PCB, or are you using some sort of prototyping board? \$\endgroup\$
    – Edgar Brown
    Commented Oct 30, 2018 at 5:26
  • \$\begingroup\$ yes layout with board.layout image also added in question. \$\endgroup\$
    – Transformer
    Commented Oct 30, 2018 at 5:34
  • 1
    \$\begingroup\$ if your scope inputs are not electrically isolated you can't measure relative to two different grounds at the same time and get sensible results. \$\endgroup\$
    – Jasen Слава Україні
    Commented Oct 30, 2018 at 6:29
  • \$\begingroup\$ Yes you are right , i break the isolation here, will try to measure again of Ch1 on other side of the isolator. \$\endgroup\$
    – Transformer
    Commented Oct 30, 2018 at 8:07
  • 2
    \$\begingroup\$ I doubt this has anything to do with the problem (RS485 is not that fast) but that ground distribution does not make sense to me. Field ground should cover all of the area under the 485 lines, signal ground should stay away from them. \$\endgroup\$
    – Edgar Brown
    Commented Oct 31, 2018 at 6:00

1 Answer 1

Reset to default
1
\$\begingroup\$

The ground layout looks wrong. The break between signal ground and field ground should run under U3 splitting it half.

Looking at the schematic, keep everything to the left of U3 over the signal ground. Keep everything to the right of U3 over the field ground. The way it shown it does not make sense.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.