Your network is the most important part of your control and process. If it is not designed, installed and certified properly, you are leaving yourself open to a wide range of problems. Many of these problems may be elusive or intermittent and can stop or interrupt your process and business, costing precious time, resources, money and, perhaps, product. Here is a simple list of steps you can follow to guide you through a new network installation or to help identify problems you may be experiencing with your current system.
COMMON PROBLEMS FOUND IN THE FIELD:
1. Trunk or spur lines should not exceed the maximum allowable distance based on specs.
2. Shields are not grounds and should not be tied to the same ground bus as the power grounds. Shields should only be grounded at one location to avoid ground loops. Multi-point grounds can be done in certain situations, but must be done correctly. It is common practice to heat-shrink any bare shield wires to avoid accidentally grounding the shield of one segment to another.
3. Proper separations should be maintained between network cables and high voltage cables.
4. Loose connections can cause intermittent issues. A loose wire at a terminal block or connector can cause sporadic communication issues with a single instrument or entire segments.
5. Segments must be designed to handle the voltage and current loads for all of the instruments on the segment. Overloading a segment with too many devices can cause devices to drop off a segment, or pull down an entire segment.
6. Instruments should be certified and designed for the type of network they are being installed on.
7. Trunk and spur cables should be checked for damage, and cables should have proper bend radius. Damaged cables and connectors can lead to short circuits and cause loss of instruments or entire segments. Exceeding the bend radius can stress cables and shorten their life.
SYMPTOMS TO LOOK FOR:
1. Loss of communication to the entire network segment or individual devices.
2. Intermittent loss of segments or devices.
3. Bad or “frozen” values in the control room from instruments.
4. Excessive LF (low frequency) or AC noise present on a network segment.
5. Devices sending error messages, re-transmits, or bad quality status.
LESSONS LEARNED FROM EXPERIENCE:
1. Verifying you have avoided the problems listed in the common problems listed above is a good start.
2. Testing a network segment before you go “live” will save both time and money. Pre-testing a network segment before instruments are connected can save precious time on trouble shooting. If you know your trunk and backbone hardware are installed properly before connecting your instruments, you can usually eliminate the trunk and its hardware as sources if you have issues once you begin connecting your instruments.
3. Certifying a network segment once you have powered up will give you the peace of mind of knowing your network segment is installed properly. Also, you will have documentation and a baseline in case you experience a problem or fault in the future. Always certify a network after ALL OF YOUR DEVICES are installed, powered up and commissioned. Certifying a network or segment before every device is connected and commissioned will NOT give you a true status of the health of your network.
4. Always get documentation! If it is not documented, it never happened. Make sure you get hard copies of ALL tests performed on your networks. You can then use this documentation as a reference should a problem arise in the future.
5. Train plant and maintenance personnel. Having plant personnel that can perform basic troubleshooting tasks can be extremely valuable. A little knowledge can go a long way.
If you have or think you are having problems with your network or installations and need advice, please Omni at 908-412-7130.
Comments