Problem with LDCM module of RED670

[DiMiTKo]

Hello! I have a lot of problems with the communication between LDCM modules of three end overhead line. The problem is that I lost receive signal from time to time from 100% to 20-30%, and I receive synchronization error. For example for 24 hours I have 10-15 comm fail signals. I use recommended settings for LDCM in the manufacture manual. I have three end so, I have 6 LDCM modules and three of them are masters, three are slaves. I switched off time synchronization to all of them, and I use Echo mode for differential synchronization. I made loop back test from the other end and i saw that some of them making problem with the communication from time to time. When I made loop back test on the terminals of the LDCM with the short cable all of them was working perfect. I'm not sure about the setting for the power of the signal. I saw that on low power they working better, on high power they generated mpore comm fail signals. The lenght of the overhead line is 17 km and we use average distance LDCM modules. Is there someone who can give me advise where could be the problem. Here on site I have one collegue who mesure the fiber optic and he told me that we have 8db optical budget. It is 2,5 less than the device expect.

 

[inpran]

Dimitko,

Why is time sync in off condition? If I recall it right, your RED670 won't work if not time synced. I'm at site right now, will confirm in some time.

But meanwhile if you get the technical reference manual of that relay, please go through it. It clearly says it won't work when the relays are not time synced.

 

[DiMiTKo]

I'm not agree with you. In their manual they don't recommend time synh source. It is not same. In the RED670 time synh and differential synh are different. If you take a look in this manual I use 5.1 combination. See the link.

 

[ptcabb]

With direct fibre connection you do not need time sync, echo is perfectly fine. GPS time sync is only done for asymetric communication paths, such as can happen when using G.703 over a telco network.

Loss of communication can be the result of settings, hardware, even firmware to some extent (if you have an old RED670, version 1.0 or 1.B you should upgrade).

I´ve never had many issues with the LDCM modules themselves, and since more then 1 link is bothering I don´t think it´s a hardware defect. That leaves the fibre, settings and firmware. Usually default settings are fine for direct fibre links, you change the low or high power at most. At 17km you should probably set it to low power, but it depends on attenuation really.

What is your firmware version? You can find it on the HMI under diagnostics. And the LDCM hardware number, should look like 1MRKxxxxx. Might as well add the serial number of the box, T.....

But this is all very vendor specific .. if you want some specific help, you can contact me at dagath at gmail.com (not posting my abb adress here ) and I´ll come back to you.

 

[DiMiTKo]

Hello ptcabb! The firmware version is 1.1. LDCM modules are for average distance, now I'm not on site but last letters was AAr02. The attenuation of the hole line is 7,5dB. For now with decided the problem with 7dB attenuators from each side. We put on the beginning of the line one 7db and for know is working pretty good, but we still watching. Actually when I put in operation RED670 with two ends I never have any problems. But always when I have three ends I have a lot of prob;ems. Can you tell me how is correct to work LDCM like master-master or mastel-slave. The topology of the connection is that each protection send to the other two his current and each protection calculating the differential current. In the manual is confusing in what mode they should work. One my collegue told me that the connector of the protection should not be tight and it is good to have some space between the fiber and module. This is very strange for me.

 

[ptcabb]

The medium range LDCM module has a standard FC connection, it's a screw in type. There's really not much you can do wrong here. Unless damaged somehow there's only one way to insert it, and the connector should make sure any alignment is ok. I wouldn't say you have to force it in, but there isn't a lot of room to manouver it either.

It's quite easy to mishandle fibre optics of course, wiping a finger over the connector and then brushing it up with a paper towel I wouldn't recommend. And of course not making any U turns with the fibres, etc. Or forcing it through some cable duct. Water damage for example can be really difficult to find out, at one point it works and the next nothing.

The master / slave principle is purely related to the LDCM link, it has no relation at all with the differential function. In fact, the differential function doesn't really care if you feed it from a direct analog input or through a LDCM, in that sense it doesn't know any communication exists (although the algorithm does have optimisation regarding the time delay introduced by any communication link).

You put one of the IEDs in master and the other in slave. The master will handle the time sync, it will act as the source. When using echo the latency will be measured on a round trip of a package divided by 2 (that is why send and receive should have the same time delay, if not you need GPS time sync).

Then you choose an address for the local and remote terminal. For example 1 and 2. On the remote side of the line, you choose exactly the opposite, 2 and 1. Additional LDCM links can in theory have the same 1 and 2 address, but it is sort of bad form - if for whatever reason the fibre optics were switched, the IED would assume it is still connected to the correct line end. So it is best to choose different addresses for multiple LDCM links, 1-2, 3-4, 5-6, etc.

OptoPower has to do with the attenuation. General guideline would be, everything under 1km = low power, everything above = high power.