Money4Nothing

DeviceNet to ControlNet

9 posts in this topic

I have a ControlNet network with six 1788-CN2DN linking devices to scan six of my DeviceNet networks. I am in the process of commissioning, and everything works great.....except....... When I power up the network (all CNet and DNet modules together), sometimes one or two of my CN2DN devices will not initialize communication with its DeviceNet network. The Module Status LED will flash red and the DNet Network status LED will also flash red. This apparently indicates a communication timeout, even though all devices are powered up simultaneously. But the rest of the networks will come online just fine. All 6 networks are identical, with 18 nodes each, and using ADR. The DNet devices are Allen-Bradley E3 overload relays. Whenever I power up, it seems like one or two of the networks will randomly fail to initialize. RSLinx and RSNetworx can both see the CNet modules, but the DNet is not online. Only by recycling power again to the CN2DN module can I get the DNet network to come online. This is a problem because the operator might not notice the communication timeout message until its too late; also, modules are in a remote location, so its difficult to recycle power to each one individually, and its undesireable to recycle power to the entire system. Besides, I don't want to ever have to recycle power, I want the doggone networks to come online at power up correctly! I'm working with Allen Bradley on the problem right now, but if anyone has any experience with CN2DN networks or any ideas, please give your suggestions. FYI these CN2DN networks are part of a larger system with 15 Control Net nodes and a Redundant Control Logix processing platform, multiple HMIs, ethernet, etc, but everything else is working fine. Thanks $

Share this post


Link to post
Share on other sites
Couple of Questions and some suggestions. Are your CN2DN modules listed in the I/o tree of logix 5000. If so rather than cycling power try a right click on the module and select reset module. If this brings that subnet up then you can programatically use the GSV to find faulted CN2DN modules and an SSV to reset them from PLC code. I found this approach to work with 1734-ACNR modules connected to 1734-Rs232 modules in a Controlnet environment.

Share this post


Link to post
Share on other sites
I have commissioned a lot of CN2DN's both with PLC-5C15 controllers and with ControlLogix. You can get the exact condition of your DeviceNet network in the Status tags that are built in the ControlLogix Controller-scope tag database. This is the same information that would be scrolling across the dot-matrix display of a 1756-DNB. CN2DN_Bridge:S.ScannerStatus CN2DN_Bridge:S.ScrollingDeviceAddress CN2DN_Bridge:S.ScrollingDeviceStatus CN2DN_Bridge:S.DeviceStatus[x] where x is the node number. It sounds like you're getting a simple Bus Off error, probably because of a power issue. What kind of power supplies are you using on your DeviceNet ? DeviceNet is fairly picky about the quality of the power on the V+ and V- lines, and it is very finicky about the "rise time" when the network powers up. Your power supply needs to be able to go from 0 to 24 volts in under 250 milliseconds, or the DeviceNet transcievers will not power up and establish the bus signal correctly. The E3+ runs it's logic board as well as the output relays on board from the DeviceNet power... they draw about 3.9 watts steadily (at 24 VDC, that's 162.5 milliamps per E3+) and they say they can draw up to 3A surge at powerup ! Maybe your power supply can provide the 2.9 A the overloads would need steady-state, but ends up powering up more slowly because of the inrush. Also, though it's very tempting, it's not strictly proper practice to provide the CN2DN main power from the same power supply you are using the run the DeviceNet. DeviceNet should have dedicated network power supplies, always.

Share this post


Link to post
Share on other sites
Thanks for the replies. The devicenet and supply power come on simultaneously, and the power supply i'm using goes to full voltage within 250 ms. Its a Sola SDN powersupply, rated for DNet. But somehow the scanner cant get a bus signal quick enough some of the time, and goes into minor fault. When I set my Transmit Retries parameter of the CN2DN to 5 or higher in RSNetworx for DNet, it seems that I have no problem connecting. However, when I recycle power to the scanner, or if I reset the module from Logix, the Transmit Retries parameter is reset back to 1. I believe this is the best solution, but I can't figure out how to keep that parameter from changing back to 1. I'm downloading the change online to the CN2DN, and I'm saving the file. I can't find anywhere to change the default setting. Any Ideas? Thanks $ Edited by Money4Nothing

Share this post


Link to post
Share on other sites
You should not have to alter the Transmit Retries value. This is a physical media problem and you should address it instead of trying to change it's effects in software. Are the power supplies for the DeviceNet and the 1788-CN2DN actually separate ? You might also have a bad power supply, even though it's one that is approved for this application. What is the scanner fault code when this happens ?

Share this post


Link to post
Share on other sites
I have tried configurations with seperate power supplies and one power supply, there is no difference in the performance. The error is a buss off condition. There is something about the power that the scanner doesn't like when it first powers up, and it doesn't establish a DNet bus connection. It doesn't seem matter if the DNet bus is the only load on the power supply or not. I have the shield and 24V- (blk) grounded at the power supply. I put a scope on the 24V DNet power. At 20 ms there is rise up to 15V which lasts about 15ms. It stays near 15V with a couple small oscillations until about 70ms elapsed, then rises up to 24V in the next 20 ms. After that its a nice steady 24V. This looks pretty normal to me. I also scoped the voltage from CAN_L and CAN_H to ground. After 7ms there is a jump to 1V, 8ms later there is a rise to 4V over the next 20ms or so. After a couple of 100mV oscillations around 4V it settles down at about 50ms elapsed. Obviously there is no traffic since CAN_L and CAN_H are equal. I'm not sure why CAN_L is so high, it should be no higher than 2.8V under normal operation. Everything looks like its wired up to spec. I'm beginning to suspect that my field devices are somehow causing a problem with the DNet power bus. This problem is the same on all 6 of my identical networks. Maybe its possible that the scanner bus connection wants to get a 24V reference faster than the power supply is coming up to full voltage. But 100ms is pretty fast, and within the DeviceNet spec. The strange thing is that sometimes one of the scanners will connect (there are 2 per enclosure, sharing a power supply), and sometimes neither will. It seems totally random. However, I can't identify any sources of noise. Maybe the problem is that there are 2 DNet power busses powered from the same power supply? I have never seen anything to prohibit this. I'm still working with Allen Bradley on the problem. They want me to try disabling ADR but I can't see how that will help. Thanks for the replies. $ Edited by Money4Nothing

Share this post


Link to post
Share on other sites
Doesn't take too long and can't hurt. How about the physical network. Cable type, terminators, trunk and drop lengths, etc?

Share this post


Link to post
Share on other sites
Wow so far I'm still baffled. Allen-Bradley hasn't come up with a solution yet either. 1 remote junction box containing two CN2DN modules. One 120V source of power to the box. Each CN2DN is a network master. Each network has the following: 18 nodes w/ ADR (E3 overload relays) 250 ft trunk, round media thicknet 9 drops of 10ft each, round media thinnet 125k baud There has got to be something fishy about the way I have it wired up. I have the same power supply for both DNet network power busses, and I thought that might be a problem. So I split them on seperate supplies. No luck with that. Disabling ADR had no effect. Terminations look good, I see 60 ohms between my data conductors. Once I'm powered up and reset all the connections, the networks perform beautifully. Yesterday I powered up the 600V motors for the first time and there were no noise problems at all, even on heavy loads. I even tried it with both networks and comm modules powered from the same 24V supply. I also put my trunk line right next to my 600V power cables for long stretches. I couldn't make it lose communication at all Next week we'll power up again and I'll increase the baud rate and do some other things to try to make it fail. Its still very strange to me that they only fail when power to the CN2DN module supply and power to the DeviceNet is applied simultaneously. If I wait a few seconds to power one after the other, there is no problem. But since my 24V supplies are in the remote Jbox, I can only switch on the 120 AC feeder back at circuit breaker panel. Its impossible to switch the 24v seperately unless i added hardware into my enclosure, which is already built and installed. I still have a few ideas. I'm going to look at my Can_H and Can_L on a scope to see if there is any 60Hz noise on them during power up. I'm also going to drop out most of my network and just concentrate on my two closest nodes. I had done that before, but with both networks powered from the same supply. This time I'll isolate them. I might also increase the size of my grounding wire. Its 16 awg right now, but its only 2 feet long. I'm gonna keep trying to eliminate possible sources of problem, but at this point I still have no clue of a cause. Any suggestions on what to try to get some more clues to the problem would be appreciated :) Thanks $ Edited by Money4Nothing

Share this post


Link to post
Share on other sites
Hello guys.

Can someone help me???

I'm having a big problem with a 1788-cn2dn module in an application. The module changed the behavior after a test I made to return the electrical connection of a GE DH485 communicator (devicenet to MODBUS RTU). Is this behavior that I am sending in the video is some serious equipment failure? Additionally, in RSlogix 5000 you have an exclamation icon in the I / O Configuration that indicates that module 16 code # 0204. Is there any Compatible Keying?

Thank you for your attention.

1788CN2DN_00.PNG

1788CN2DN_01.PNG

1788CN2DN_02.PNG

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now