H2-ECOM/ECOM100 problem

[Gummi Ben]

Hi, I have a system of PLC all connected to a ethernet network, some PLC's have H4-ECOM and some have H2-ECOM and H2-ECOM100. They are all communicating with a SCADA system via OPC server. Sometimes it happens after a power loss that the communication to the H2_ECOM / ECOM100 modules stops working. If I try to edit a link to the PLC in DS I get a error messages. The modules show up in NetEdit and if I update the firmware to its current version ( no change) it starts to work again. I wonder what might be the fault and even why this "fix" is working ? Could it be that by updating the firmware, I'm resetting the module. I tried to power-cycle the PLC but that was not enough. Any thoughts ? I'm not having this problem with 5 other modules that are H4-ECOM or 2 pcs of H0-ECOM.

Best regards,

Gudmundur Ben

[Greg]

Hi Gummi Ben,
If the ECOM/ECOM100s are initiating the sending of data to a SCADA system (i.e. the ECOM/ECOM100s are masters and the SCADA is slave), then it could be a programming issue or communication across the backplane. One of the things that plague DL405 systems is the connection to the backplane. Make sure the H4-ECOM is seated properly in the base and does not have any bent pins.

Also, by updating the firmware (even with no change) you are indeed resetting the module and rebooting it. But the fact that a power cycle doesn't clear this may mean something else.

To see what is really happening it will be necessary for more details. Specifically:

• 1. Is the SCADA polling the ECOM/ECOM100s? Or the ECOM/ECOM100s initiating the send?
• 2. If the ECOM/ECOM100s are initiating the sends, then we may need to troubleshoot your program. So any details you can provide may be helpful.
• 3. When the ECOM/ECOM100s are not communicating, what is the status of all the lights on the front?
• 4. Is the "power loss" a voltage spike? Or does it just go down for a long time?
• 5. Is it possible to get a Wireshark trace of a "good" network and then a "bad" network?

BTW, Wireshark is a free downloadable Ethernet packet sniffer. You can install it on your SCADA PC and should be able to see all the communication happening between the ECOM/ECOM100s and the SCADA.

[Controls Guy]

If I understood him correctly, his 0's and 4's work fine and the problem only occurs on the 2's.

[rlp122]

I think one more question is in order. 

What is "the current version" of the firmware Gummi is uploading?

[Gummi Ben]

Hi and thanks for all your answers.  The SCADA system is polling but the PLC's are also sending information one to another.  I don't think the problem is in the PLC code since it works well most of the time, and this problem only occurs after a power failure and most likely a voltage spike (it's not off for a long time).  I'm not sure what the status lights are when it's not communicating, but I will look at that if this happens again.  This has happened few times over the last 2 or 3 years.  This is only a problem there with the 2's as Controls Guy pointed out and I dont think the firmwares versions are important because I have updated them and this has happened with more than one version of the firmware. 

[rlp122]

The firmware versions are important, See this note on the NetEdit3 utility:

"Note: As of version 3.12, earlier versions of NetEdit3 will no longer be able to detect and download updated firmware files. It is highly recommended that you upgrade to the latest version of NetEdit3."

You cannot be assured that your versions of firmware are current if you are not running NetEdit3 v3.12.

[Greg]

Gummi Ben it is impossible to predict the state of electronic components after a voltage spike. Every precaution is taken to try to insure things shutdown and power up in an orderly manner, but a brown-out and/or spike can cause indeterminable modes. I take it that after this weird condition that a normal power cycle clears things up? If so, I'm not sure we can make the H2-ECOM100 completely immune to what I would call "dirty power." So, I would recommend you install hardware that will maximize the isolation of this PLC system from brown-outs and spikes.