Emerson Exchange 365

In reply to RickV:

Thank you RickV for your response. 

Greatly appreciate your help as I been chasing this noise for a few weeks now. Learning alot in the process. 

Please see my comments in red

RickV said:

AO Cards have isolated channels. I suspect that you are getting noise from the field and or wiring. Are the wires for the channels in the same conduit or follow the same path by chance? (I will confirm if these wires for both channels are in the same cable route)  I would start with looking at the following:
1) Verify that the field wiring is twisted pair shielded cable. Confirmed twisted pair shielded cable
2) The shield should be connected only at the DCS side (shield bar on carrier) and not in the field at the valves Confirmed.
3) Shield bars should be grounded to good earth ground in the DCS cabinet then to good plant earth ground. The carriers are jumpered together and the last carrier on bottom is wired (with #16 awg wire) to cabinet frame enclosure. We have another cabinet (in the same row of cabinets) with #1/0 awg terminated to an enclosure frame ground lug, and that cable is terminated about 25 feet away at a Local Master Cabinet ground bar (with other Cabinet grounds), and then from there another 1/0 cable ran to terminate a plant earth ground.
4) Is there marshalling terminals, and if so are the shields carried through properly Confirmed.
5) Verify that no AC is running adjacent with any of the wires Does this mean to look in the cable tray where these wires are located and makes sure only instrument cables is in the cable tray and no AC wires is in these cable tray. AC wires meaning motors, VFD etc?

In reply to RickV:

RickV,

Just some background for this investigation. The valve on Channel 2 has been causing intermittenly No Communicaiton Error (losing the HART). The valve was previously on another AO card and was experiencing intermittenly I/O failures causing the valve to lose signal and go to fail close. This is the 3rd positioner on that AO card and channel. So we moved to different AO card thinking that would cure the issue, but now we see still see  No Communicaiton Error (losing the HART). Lots of HART timeout errors as well.

The problem is due to that noise.

Do you think these high voltage spikes will eventually damage the channel?

Thank you

In reply to DCS Newbie:

Eventhough the AO card have isolated channels, it does share one HART modem. You may see other channels having intermitent HART timeout error as well can causes HART communication errors on other channels as you indicated. I have seen where the device could the be root cause to these noise issues you are seeing. Double check the Valve before you start the costly rewiring. BTW: to your question on RickV question 5, No AC wiring should be with any DC wiring especially with any bus communication. Should aways be on seperate wire tray.

In reply to Tinh Phan:

Thank you Tinh Phan for responding. 

I was not aware the AO card shares the HART modem with all channels. Are you stating that the noise from Hart enabled channel can cause communicatioin errors on another Hart enabled channel?

For this AO card, the HART channel 3 thru 8 has been disabled.Only channel 1 and 2 has the HART enabled.  When I measure the noise on the other (3 thru 8) channels, I see low noise (less than 400mVpp). When I disconnected the wires from channel 2, I measure low voltage (120mVpp) directly on channel +/- terminals. I measured the noise on the wire when removed from channel 2, and it was 2-3Vpp. I was unable to disconnect the wires on channel 1 and measue the noise on the wire cause that valve is critical to operations.

If the HART is disabled on the other channels, 3 thru 8, can I still see noise on those channels 4-20mA signal?  

This is interesting because the valve (Fisher) on channel 1 has 0 Hart Timeout Errors when monitored in Diagnostics. The valve (Flowserve) on channel 2 has about 40,000 Hart Timeout Errors when monitored over a 6 day period. For channel 2, I also get Failed to Read Device Alerts every 5-10 minutes.

Flowserve Valve specialist has come out here and checked the valve (channel 2) and stated is working normally. I can ask Flowserve to come back out and hook up there HART communicator (laptop) directly to the valve and monitor for Hart Timeout Errors. Valve on chanel 1 is currently being checked for shield not connected at valve.

Thank you for your help

In reply to DCS Newbie:

I am not stating that it will happen but it could especially if there is large enough noise spike but nothing like what you indicating Fisher vs. Flowserve. You may see random Hart timeout error in small packages when the noise is high but the HART communication is still good. If your Flowserve having that many timeout and other HART failures, you need to focus on that valve electronic and the wiring.

As for the Flowserve indicating the valve is working fine, did he means just the 4-20 mA or both 4-20 mA and HART communication?

When you are buying HART instrumentations, I recommend that you verfiy with the HART foundation to see if that instrumentation fw, rev, etc has pass the HART certification. This will gives you a comfort that the device manufacturer is following HART foundation spec.

In reply to Tinh Phan:

Newbie,

Just noticed something on your first responses in red. You mentioned that all the shields are jumped together and the “last carrier on bottom is wired (with #16 awg wire) to cabinet frame enclosure”. If these are Horizontal carriers I like to see each shield bar before an extension cable connected to the ground bar. If these are vertical carriers you’re ok.