Control Logix Rack Card Faulty Connection

Hello, I am having an issue with a control logix rack we use in a conveyor system. We have a DH+/RIO card in the rack used to bridge to a data highway network used in a remote area. The problem is that when i connect to the rack in Linx through an ENBT/A card, the dh+ card keeps losing connection (the red x appears on the card then disappears then appears again at irregular intervals.) I have changed the DH+ Card with a card that i was able to maintain communications with in a similar test and i have also changed the chassis because i have had individual slots go bad after receiving electrical spikes. So I Guess i'm wondering if it could be the ENBT/A cards Backplane.
ControlLogix Configuration
Slot Module
0 1756-L61 5561 Processor
1 1756-IA16I/A
2 1756-IA16I/A
3 1756-IA16I/A
4 1756-OX8I/A
5 1756-OX8I/A
6 1756-DHRIO/C This Card Falls in and Out
7 1756-ENBT/A Used for Control
8 1756-ENBT/A This is the ethernet card used to connect.
9 1756-CNB/E

Is it just an RSLinx RSWho "connection failure" animation(node red x-ed) when browsing the network or you are experiencing DHRIO connected networks data transfer failures?
What about the LEDs on the front of the 1756-DHRIO module's state?

When navigating through rswho to the ethernet card i can see all cards from the enbt backplane except the DHRIO card is intermittently displaying the red x animation in rswho. the front panel indicators are all flashing green (chan a, chan b, and the OK light) I have been reviewing and altering the routing table but it is the same routing table and slot configuration i have used on other currently running and functioning data highway networks. there are transfer failures of the dh network and we are checking connections and changing out cables to correct this; however, this shouldn't affect the visibility of the DH card through the enbt\a backplane should it?

Is there any alphanumeric code (other than "normal operation":
A DH, A#xy, A OK) accasionally flashing on the module's display?
The 1756 backplane has plenty of bandwith for the RSLinx utility to "notice" a properly functional DHRIO module.
The fact that you are having data transfer issues over the DH+ network bridged by the 1756-DHRIO module and RSLinx "occasional invisibility" issues kinda points to hardware and/or DH+ media issues.
Do you have any SCADA application communications routed through the DHRIO bridge?

This might also help:

http://literature.rockwellautomation.com/idc/groups/literature/documents/um/1756-um514_-en-p.pdf

Just a thought....

...I've recently seen similar issues when any one I/O card is set for the fastest (0.2mS) RPI setting. It seems to hog backplane comms, leaving RSLinx fighting to get a word in.

Also if you are not using COS to trigger event tasks, then disable COS On-to-Off and Off-to-On in the input modules.