Our neighbor's 13 year old DeltaV system had a serious fault yesterday when the the first connector in the 24 VDC I/O power daisy chain became a fuse - it melted, opening up the circuit and interrupting power to the 5 downstream 8-wide carriers. The individual wires were maybe 16 AWG and the ones closest to the fault were "browned" like they had gotten a little warm. When the 8 terminal connector was pulled from the leftmost position of the first backplane, part of the backplane came with it. He was able to terminate and insert a spare that came from one of the other backplanes, which restored power once fully inserted.
Neither of use could locate a spare 8-wide carrier so we have so far not had a chance to remove it and inspect it. There was some urgency to get the plant up and running in the cold weather.
A clamp-on ammeter showed 4 amps immediately upstream and downstream of the suspect termination. The user added a pair of conductors to connect the "last" 8-wide carrier in the series directly to the power supply, so the six are now fed from "both ends" if you will.
The facility is normally unmanned and no one was in the locked control room when the fault occurred. There was no recent work done, and no signs of dust or water ingress. The first 8-wide carrier (the one with the fault) has only 24 VDC discrete dry-contact input cards. The next 8-wide has 8 24 VDC DO cards, all wired to interposing solid-state relays. The next few 8-wide carriers have analog IO and serial cards.
1.) Anyone ever seen an issue like this before? What caused it?
2.) What are typical readings for current draw?
3.) Any recommended AWG for the conductors connecting 24 V power to 8-wide carriers? Ours appear to be a little bigger.
4.) Any problem to feed them from "both ends"? The original integrator didn't do this, and I noticed ours isn't fed from both ends either.
Your thoughts and comments are welcome. We have never stocked an M-series 8-wide carrier as a spare part but I think we will be now.
John,
We had an issue on an M-Series DeltaV system 8-wide I/O carrier in May 2009. This system consists of three 8 wide I/O carriers. The first carrier has the Pass through Power Supplies & Controllers (Primary/secondary), the second and the third 8-wide I/O carriers have Analog and Discrete I/O modules.
When the problem started, all Field Instruments (connected to all three 8-wide I/O carriers) went to their failed states (IOF on every channel) and the unit shut down. After trouble shooting we were able to isolate the problem just to the third 8-wide I/O carrier (when the connector cable from the second carrier to the third carrier was disconnected, all the instruments connected to the first and the second carrier could return back to normal operation). Further trouble shooting indicated the connector cable to be O.K. and the left most connector block on the third I/O carrier (onto which the cable from the second carrier is connected) was found to be faulty. We got a replacement of this connector block from Emerson and restored normal operation to the third I/O carrier.
Unlike the case that you mention we had no visible indications of melting or burning, either on the cable pins or on the extender block. We have several DeltaV systems at our facility and this was the only instance of seeing a failure on a I/O carrier sub-system. Our systems are not fed from both ends either. The 8-wide I/O carrier is the only common feature. We don't stock any spares either.
Thanks
Kris Chandrashekar
Here's a few thoughts from what little I know.
The power connections on the top of the DeltaV I/O carriers provide power to the field devices, which means that the current through the "daisy chain" will depend on what current is drawn by the field devices.
The Product Data Sheet for the M-Series Horizontal I/O Carrier specifies a maximum of 6.5 mA for the field power connection, which includes the total current draw from the devices connected to all of the carriers daisy chained together.
Below is an analysis of the maximum current levels possible, given the first four carriers are each fully populated in order by DI Dry Contact, DO High Side, AI and AO cards, all 8 channel cards.
The specifications on the 24 VDC DI Dry Contact card provides a possible maximumof 40 mA per card, for a total of 320 mA field power current draw for a carrier with 8 DI Dry Contact cards installed.
Assuming that the 24 VDC DO cards on the 2nd carrier are High Side DO cards, the max current for the field power can be 3 A, depending on what the devices are drawing at that point in time.
The AI and AO cards can have 64 channels for each carrier - assuming that the AI cards are 8 channel and not the 16 channel version. Using the 20 mA of the 4 - 20 mA loop as the maximum current possible (could actually be 22 mA each channel), the current draw for the 64 channels could be as much as 1280 mA.
The current draw for the first 4 cards could be as high as 5.9 A.
With the first four carriers - assuming the card layout as I described, has a maximum total nearly to the I/O carrier maximum specification.
Since the wires to the first carrier showed signs of heat, the wires had to have been carrying current close to the limits. The currency capacity of wire depends on several factors, including the length of the wire,the gauge, the wire type, the allowable voltage drop, etc. For a wire sizing procedure, check www.sunforceproducts.com/.../How%20to%20Size%20Wiring%20for%20Your%20Battery%20System.pdf.
The new connection to the last I/O carrier can be useful, but may not solve the issue, and may cause additional problems. A ground loop has been created in the field power connections and could cause issues with the signals to the devices. Since that new connection is in place, the daisy-chain for the field power for the six carriers can be separated to three carriers for each power connection by disconnecting the jumpers between carrier 3 and 4 ( or between carriers 2 and 3, depending on an analysis of the field devices and expected current draw). That disconnection could result in a transient "variation" in the field signals, so the disconnection should be done with the process in a state that could absorb any such variation (process is shutdown?), or with plenty of operfational assistance to help overcome any such variation.
The site may be lucky that the connector failed; at that current draw through the first carrier, components on the carrier could have failed (the basis for the 6.5 A limit) and a new carrier would have been required.
Good luck!
J.D.
In reply to JDWheelis:
My third line should read:
The Product Data Sheet for the M-Series Horizontal I/O Carrier specifies a maximum of 6.5 A for the field power connection, which includes the total current draw from the devices connected to all of the carriers daisy chained together.
The entire post should be: