NI-3000: It works, but I can't communicate with it.
deepforestgreen
Posts: 51
I have an NI-3000 controller (I know, old) that does its job until I need to communicate with it. For the first timei n years we need to change the programming, and I can't get a rs-232 connection. I've adjusted the dip switches to try every baud rate and I've used two different cables (both are often used for this exact job) but nothing communicates. Any ideas on what I haven't tried?
0
Comments
You've but probably already tried this, but in troubleshooting you can't assume anything.
Use a crossover cable with a static address on your laptop. Of course this requires knowing the IP address on the master already. And knowing that the master is not set to that damnable default of DHCP. If it's not connected to a DHCP enabled network, then it takes an extra year for the master to reboot while it times out its DHCP attempt. Or connect to the programming port and get/set the address. Nice catch-22 if you can' get serial to work, eh?
If you know the non-network-connected master is not set to DHCP and has an unknown static address assigned.. then your static address and crossover cable can probably find it by using "Listen". But that's a lot of unlikely ifs.
Good advice, but I have tried all of that. Thanks, anyway.
This is exactly what I just did. "Listening" didn't catch anything.
I know the IP address it was supposedly given five years ago. I can't ping it. When I plug into the same router (right next to it) and give myself the same subnet, etc., and try to ping it or "Listen" for it then nothing is found. I am simply one IP address away on the same network (on the same router) and I'm picking up nothing.
The rs-232 PROGRAM port is not responding. I notice that one of the dip switches (#4) doesn't stay in either position well. I've tried every baud rate combination, however, and nothing is working.
Can you get anything IP/TCP to work? FTP or telnet?
I can't even ping the thing. I suspect it has an IP address that the owners have forgotten. I have no way to know it.
Hedberg has the next correct suggestion. Set the dipswitch that disables the program and reboot it. Then try to connect.
There's a great app for the iPad called Ping Lite and it has an option to ping the subnet and it will show you evey IP in that subnet that responds to the ping broadcast.
It's possible that UDP BC rate may have been set to "0" in which case "listen" may or may not work. Others might know for sure. One of the AMXers explained it a years or so ago and I don't recall what was said.
Check the router's DHCP table or DHCP "static" binding assignments.
Done that. No effect.
Definitely running out of suggestions.
The iPad app "iNet" is even better, it shows the name and services any device found is offering. Fast and very nice, the free version was recently downgraded to help sales of the premuim one, but the price is low and the features are high in the pay-for version, so I don't mind. Much.
Yes, if UDP BC is 0, no beacon, nothing to "listen" for. We turn it off to reduce pointless network chatter, as various AMX and other network items constantly respond to these 5-second default beacons. The downside is making it harder to find.
If a network scan can't find it, i think you own a brick there.
Here's iNet showing part of my network with the 3100 netlinx at 201, a 3000 at 210, a computer and some panels in the 190's. I have found this invaluable on several occasions.
1) Connect to the console at a known baud rate and hit [enter] a couple of times followed by reboot[enter]. If the master reboots, it probably has a dead capacitor that is part of the charge pump for the RS232 transmit interfaces and is working correctly but is unable to respond to you. This is a known issue that is discussed in other threads.
2) Connect a crossover cable between the Ethernet interface on the NI-3000 and your laptop. You probably won't need a crossover cable as most (PC) NIC's autonegotiate these days, but you'll know if you get link lights on the NIC's anyway. Get a copy of Wireshark and watch what packets flow across the wire. You can generally determine the IP address of the master by examining where the traffic is coming from.
oo, that's clever. Hadn't thought of that.
The problem that I'm having with the listen and with iNet (or other program) is that you probably need to know what subnet you are on for them to work. There are a lot of subnets to try. Will Wireshark pick up and identify stuff that's not on the same subnet as the computer's adapter?
Hooking up to the unit directly is the best bet to make sure all packets from the device hit the NIC on your machine. Using an old network hub (as opposed to a switch) will also make sure you get all packets if you need to see the device communicating with other devices.
Ethernet switches tend to stop traffic reaching your PC as they route traffic to intended recipients as opposed to spamming it out to all interfaces like the old hubs did.
Kevin D>
I'm going to try this, today. Will let you know.
Good question. No, there's nothing using IP at all on this system. Each ssytem we use is self-contained to be useable when the network is completely down.
It´s my guess that the 3000 has dhcp turned on so the IP address would be 169.x.x.x something. That is based on your program didn´t have a ip address used for communication if your first try to communicate is through RS-232
I have a little dhcp server installed for these situations, http://www.dhcpserver.de/dhcpsrv.htm and of course wireshark.
In the last two weeks I have had the same problem with two controllers.. both 3000 . A capasitor for the RS-232 ports "dried out". The customers called complaining that some devices stopped working. I saw LEDS for the serialports blinking red and some yellow, after a restart the yellows would disappear one by one. I could communicate through Ethernet to the master so that wasn´t affected, but the Program port was. Slowly the other serial ports stopped working. Called a colleague and he searched for a solution on the web and voilá, it was a common problem. Two hours later the capasitor had been replaced an the system worked fine. The latter master was just a copy paste.
But, I'm pretty sure that I was wrong about the other serial ports - I think they are ALL dead like some of you have mentioned. I didn't realize the capacitors were something I could open up and replace. I'll need to do that or buy another unit.
Thank you all so much for your help. You really bailed me out.
Interestingly, I seem to be finding that as the capacitor goes bad some devices seem to be affected before others. It seems that some devices (like Mitsubishi LCDs) are less tolerant to voltage variations on their 232 port than others and stop working early on while others will keep responding to the NetLinx for longer.