lynnl
Posts:
236
From:
Minnesota USA
Registered:
May 12, 2005
|
|
Re: Using digi one realport IA for communication to ethernet
Posted:
Oct 21, 2009 1:13 PM
|
|
The Digi One Realport IA is a pretty old product (that's not the problem, other than I've never seen a physical example of that model).
I can only comment on how the newer Digi One IA or IAP work.
Since the Digi's ground is isolated/floating, you MIGHT require a signal-ground connection from pin 5 of the Digi to whatever Rosemount 1420 treats as ground. This is because (and you can read the EIA-485 spec if you doubt me) a 2-wire RS-485 needs some form of "return path" either directly by wire or indirectly through shared power grounds. Since the Digi isolates the DC supply, there is NO shared power ground.
Also for 2-wire RS-485 to work, one of the two devices must have a "line bias" enabled, or there will be too much noise when neither device is transmitting. On the Digi, line bias is only on when DIP # 4 (the terminator) is ON. So although termination won't be required on a short cable, the line bias is critical on any length. If the Rosemount 1420 supplies line-bias, then the DIP #4 on the digi can be on or off.
As for the A/B, unfortunately the EIA-485 standard defines these as logically not voltage, so one can never out guess if A means + or -.
I normally find B = + (so screw 3 to B) and A = - (do screw 4 to A), which is backwards from how you are trying. On the Digi One IAP, one doesn't need to link 3&6 or 4&7, so I'd guess on the older Digi One Realport IA you also don't need to. It is only the Txd+ and Txd- which are active during RS-485 2-wire mode - the Rxd+/Rxd- will be tri-stated.
Are you enabling a Modbus/TCP to Modbus/RTU bridge within the Digi product? If so it fixes up any timing issues introduced by Ethernet encap. Modbus/RTU can be very fussy about tiny time-gaps, so if you are just using Digi Realport driver on Windows or raw TCP/IP encap of Modbus/RTU, it is possible timing gaps are causing message discard at either end.
|
|