Noise in Honeywell's DCS System

R

Thread Starter

Rahul Sharma

We have Honeywell's TDC3000 DCS in our plant... Recently we have observed too many noise counts in the UCN... The counts are observed in all the nodes.... We have in all 08 UCN nodes... (9,10,15,16,17,18,19,20 resply)... The Ground resistance is around 1.6 Ohms... All UCN taps are properly tightened... Cables are properly fastened to the LM modules of PLC... We isolated on LM which was giving a higher noise count rate and switched off the LM completely... Despite that the noise counts are increasing unabated... If anyone has come across such a problem, may please share information...

Thanks and regards
Rahul
 
W

Wayne Shimanis

Mr. Sharma:

Grounding systems can be the cause of many noise problems, especially for systems that incorporate microprocessor based control systems and instrumentation. Some of the latest design systems require that equipment grounds never see transients above 0.3 volts. In many cases these transients will cause unstable performance, malfunction, or outright damage to sensitive circuits. Some relatively recent studies indicate that as high as 80% of transients enter equipment via ground termination's, even though the grounds appear to be good.

You might consider ground transient terminators. Check www.vortextek.com

Regards,
Wayne Shimanis
 
J

John Capstack

I have seen this problem on several sites:
Once the noise was due the location of the UCN cables relative to a voltage source and we needed to shield the UCN cables with stell wire armour - ground the shielding. We eventually also place additional electrostatic shielding around the cable.

Have you added any new devices onto the plant which could be causing the noise? Often the noise appears near the noise source but not at the actual location.

Is the noise on both A & B cables? If not then it could be due to faulty cards/taps/cables. You'll have to try isolating the various complnnts. It could also be at the NIM.
 
Yes, we do have noise on both the cables... Plus the noise pattern is pretty irregular... Sometimes we have noise counts on all the nodes and sometimes it's localized... But more often than not, it's on all the nodes... We isolated a LM Module we suspected to be the source of noise... and there was no problem for sometime... But again the same irregular pattern of noise counts have started.... Earlier it used to rise at the rate of exactly 15 counts per sec... But now even that regularity is missing and has become more random... Well we also haven't added any new node or component to the system... The configuration was set up three years back, at the time of commissioning and it still runs with all the original configuration... Once earlier we noticed the noise counts (about nine months back) and we could solve it just by tightening the UCN cables... Would you please also confirm if the ground resistance of 1.6-1.8 ohms is ok..?? Is there any ideal value for the ground resistance...??

Thanks for your inputs..

au revoir
Rahul
 
We've had some UCN noise problems, usually after a lightning storm. The main problem has been failed terminating resistors on UCN taps or trunk connectors.

We had one instance where we could not identify the problem, so we rented a carrier band network testing device (CBM-1) from Relcom that helped identify the source of the noise. It connects to a pair of spare taps on the UCN and displays several values.

One reading shows reflection signal amplitude and distance. By checking the distance value from 2 locations on the UCN, you can pinpoint the location of a bad resister, cable or drop tap...

- John
 
May be UCN resistance, it should be 75 Ohms, please check every one of them ,it should be 74.5-75.5 Ohms; I have meet this trouble;

sorry for my poor english
 
P

Phil Corso, PE

Responding to Sharma's dilemma:
Check to see if a neighboring plant recently installed a large VFD system.

This situation occurred here in the USA when a chem plant installed a 3 or 4,000 Hp VFD. Subsequently, there were random trips of equipment in a refinery about 1.5 miles away. Both plants were fed from a common 138 kV service.

The fix was simple enough... reduce the "spikes" that were coupled into the common service. Note: two levels of transformation were involved,
i.e., 138/15 kV & 15/4.6 kV.

Regards,
Phil Corso, PE {Boca Raton, FL, USA}
[[email protected]] ([email protected])
 
John, Could you please give some more details about the Carrier Band Network Testing Device (CBM-1)...?? Where could it be procured from and roughly how much would it cost...??

Thanks and regards
Rahul
 
J

John Capstack

The recommended grounding level is a resistance to true earth of one ohm or less. Have any external electrical devices been installed on the plant recently, HV Motors, etc?
 
Rahul:

One source of noise is dirty plus moisture buildup inside the connectors. Disconnect the UCN connectors and clean thoroughly with a suitable electric contact cleaner. The cleaner must not have any traces of water. I suggest isopropyl alcohol. After cleaning allow to dry completely and reconnect. You must clean all connections, on a one by one basis.
 
please check the tightness of ucn cables at all nodes and also check the termination resistances at all taps.
 
F

Fabian Verdun

another one, please clean all the connector with Isopropyl alcohol and make sure don't use some home-made cable
 
Top