Can't ping all devices back at plant from remote radio link

[Amish]

Hello All, I have a master to multi-point Ethernet radio setup to monitor three remote sites from our plant. All the radios are the now discontinued Data-Linc SRM7210E 900 MHz model. I'm in the process of commissioning the third remote site. All the RF parameters look good per the instruction manual, but from the remote site, I can only successfully ping about nine PLCs back at the plant, all of which are SLC 5/05 PLCs, 16k and 32k models. That seems to be the only correlation I can find. It is very strange to me. I can't ping any of my servers, switches, ControlLogix PLCs or any of the several other 5/05 PLCs on the network. I plugged my laptop into the radio and ran a Wireshark capture from the remote site. When I ping one of the PLCs at the plant that will respond, everything goes fine and Wireshark shows the expected ICMP packets and to/from IP addresses. When I ping one of the PLCs at the plant that will not respond, Wireshark shows UDP packets and the destination IP is something different from the address I was pinging. From the plant, I can't ping the 1756-ENBT card at the remote site. I plugged in a laptop at the remote site and couldn't ping that either. I have no problem pinging the two other remote sites from the plant. One thing to note is that the remote radio I'm using at the new site was factory configured to communicate to the master through a repeater. However, I'm not using the repeater and configured the remote to communicate directly to the master. Wondering if I missed something here? I'll run this by the factory tomorrow. One thing they did tell me is that since I can ping the nine PLCs, the radio link is good since traffic is flowing both directions. Any ideas? Thanks.

[Amish]

I did get this figured out. The radio modems I was trying to use were factory configured for a repeater-to-master and remote-to-repeater setup. Though I reconfigured the software settings, there is a set of DIP switches inside that also needs to be changed for remote-to-master operation.

[Ken Roach]

Thanks for posting the update ! It's an article of faith for me that if a radio vendor says "there's nothing wrong with the radios" and presents some bullshit minimum connectivity proof like PING or a simple ASCII block test, then the radios are going to get all of my attention.