0

I'm having throubles with communication over RS485 in harsh industrial environment. We perform test in EMC lab, acording IEC 61000-4-4 = Fast transient (also known as Burst).

We used burst generator and capacitive coupling 1m long around RS485 cables. We used pulses +/- 500V.

Here is our RS485 schematic. It use galvanic isolator TI ISO1412.

enter image description here

I measured with a scope A and B lines on the RS485 connector during Burst testing. I see that there are disturbtions on A and B lines and on A-B math as well. So noise is not suppres that diference reciever, because there is some ofset in voltage level on A and B line.

See scope of measurement with disturbtion: Green = Math A-B Red = A Blue = B

PS: maybe A and B chanels are swapped - i don't know right now.

enter image description here

Can anybody help me suppres that disturbtion on RS485 ?.

So far I tried: 1) Replace 0R serial resistors in A and B line with ferite beads 2) I tried remove capacitor between Isolated Ground and Shield 3) I tried remove failsafe resistors R271 and R270, because they should be integrated in ISO1412 4) I twisted A and B wires

Nothink from above list help me solve the issue.

Any others ideas? Please...

Datasheet for ISO1412

mavit
  • 121
  • 1
  • 5
  • 1
    Did you try shorting the scope probe and tip to each other and seeing what signal you got? – Andy aka Dec 17 '19 at 18:06
  • If I'm reading the 'scope correctly, looks like your getting about +/- 40 V common mode voltage induced, but also about +/- 10 V of diff mode (green trace). Is that correct? – SteveSh Dec 17 '19 at 18:06
  • "1m long around RS485 cables". Are your wires twisted and/or shielded? – SteveSh Dec 17 '19 at 18:07
  • wires are twisted, but not shielded. – mavit Dec 17 '19 at 18:21
  • SteveSh Yes, You are reading scope correct. – mavit Dec 17 '19 at 18:22
  • What's on the other end of your '485 cable? Is it open, or being driven by something? – SteveSh Dec 17 '19 at 20:42
  • 1
    There has to be something there, some imbalance between the 2 '485 lines to cause the differential voltage to be generated. I just don't see it in the schematic. I was going to suggest removing the fail safe network, but I see you already did that. – SteveSh Dec 17 '19 at 20:43
  • You said the twisted pairs in the cable aren't shielded? Then what are the connections labeled "SHIELD" on the J2 connector, that look to be connected to chassis? And what does R11, off the J2 connector do? – SteveSh Dec 17 '19 at 21:32
  • Shield on J2 means metal chassis of our computer. R11 is somethink like resistor for reducing ground loop current. I think like here in Figure 9c - http://www.ti.com/lit/an/slla272c/slla272c.pdf in picture – mavit Dec 17 '19 at 21:41

0 Answers0