Omron slice AD2

[VAU_RMA 3]

Hi, I have encountered a strange problem. I have a profinet PNT21 IO device with AD2 slice cards. The AD2 cards has been set up with 2 channels, both with 4-20mA. There is mounted a jumper between A5 and A6, and B5 and B6. The strange thing is that when I have connected one transmitter (4-20mA) to each of the channels (ch0 and ch1), I only get a normal value on ch0 (0-6000). Ch1 gives 32767 (wirebreak). I have tried to connect a loop calibrator to each of the channels, with the result of a broken loop on ch1. This is the same on every AD2 card I have checked... I have read out the settings of the AD2 cards, and reed that it is set up to use 2 channels, both with 4-20mA (though it should give me a loop even though the software is not correct???) I have changed to a new AD2 card -> did not help. I have changed the PNT21 IO device -> did not help. I have changed the connection module -> did not help (I have also measured that it works, and measured the jumpers). Note: I have seen this on at least 2 of my IO devices. An even more strange thing is that when I tested it on another IO device with an AD2 card that both the channels works on, the loop calibrator did not manage to get a loop on ch1 there either (it did on ch0). Has anyone any clue on what the problem may be caused by, and any possible fix? Edited 18 Feb 2011 by VAU_RMA

[PdL 71]

Can you post a wiring diagram how you have connected your inputs?

[VAU_RMA 3]

Note: The loop calibrator is replaced with a MTL4114 module normally... Edited 18 Feb 2011 by VAU_RMA

[PdL 71]

OK, wiring seems correct. If the loop is not established the problem must be electrical, regardless of range setting. Do you have the AD2 wired up fixed to some terminals where you test your connections or do you connect/disconnect directly on the AD2? One thing I have experienced with the spring clamp terminals on the slice I/O, is that if you push the wire in too deep, the clamp is on the wire isolation and not the copper. You can check this by taking out the wire and check if you see a clamp print on the isolation. Another thing to check is if you both A5/A6 and B5/B6 read 120 Ohms.

[VAU_RMA 3]

I have used several "terminal blocks" when testing this, all with the same result. I use the testing probes from the loop calibrator directly into the terminals, so wire isolation should not be any problem. I have measured the jumper between B5 and B6 while connected (measured on the internal points inside / under the "terminal block". I have not measured the resistance between terminals B5 and B6, but I have tried to switch with another AD2 card that does work on ch1 (though it did not work on ch1 when mounted at this node / location). The only this I have not checked is the "bus block" (lowest part of the total slice module, transfer the bus line to all slice cards). But I cannot imagine there could be any problem with this, as I have the same problem on two IO nodes. Also I assume the signals required for ch0 (from the bus block) is the same as for ch1?

[PdL 71]

I think it is a matter of trial and error, try to find out if the problem it tied to physical location? But I agree it doesn't make much sense.

[VAU_RMA 3]

I finally found the problem... I just got a new ProcessMeter from ELMA (ELMA 7287), and it is the first time I have used it. What I have discovered is that the testing pins/cables from ELMA, when inserted in terminals on the terminal block for the ch1 side, bends the contacts inside the terminal block, doing so that the terminals have no contact to the circuit board. (This does not happen when inserted in terminals for ch0.) I have also tested with some FLUKE testing pins/cables, and they do not bend the contacts in the terminal block. Note: The transmitter did not work because I first had tried the input with the loop calibrator, thus bending the contacts so that it did not work when I connected back the transmitter. Edited 19 Feb 2011 by VAU_RMA