Fieldbus Cable Tester

[CHRISC1024]

I have a quick question on test equiptment. We have test equiptment to test a segment electically but what we've never found is a tester that can shoot the cable lenghts. Allen Bradley makes a tester (1788-MCHKR) that can be used to test several types of cables and tell you if they are wire correctly and also give you a lenght of the cable.

When certifying a network, it is usually required to document the drop lenghts as well as the trunk lenghts to ensure that the overall does not exceed the allowable limit for the type of cable being used. For the trunk, this has to be done by walking down the segment and estimating the distances between taps based on the installation. This is not the best way to perform this task but as far as I know there is no tester that can be used to determine cable lenght for Fieldbus cables.

Has anyone ever used or know of a tester to perform this activity?

Thanks,
Chris

[Mike ONeill]

Surely, the allowable limit is defined in terms of communications error. If there aren't any errors in H1 comms, the lengths are OK and measurement is pointless.
Measuring cable length electrically is usually done by resistance, but setting up for that (looping the end, disconnecting devices, etc) is probably more time consuming than walking and estimating.

[CHRISC1024]

Mike,
The problem is, as Fieldbus is becoming included frequently in new projects, usually part of the required documentation for the segment loop sheets/certification is to document both he trunk and spur lenghts. This is normally the standard for all Fieldbus and DeviceNet networks for the projects we work on, especially in the pharma industry.

On the latest project we bid, the segment loop sheets actually contain a table of all spurs and trunk sections and it is a requirement to fill in the lenghts. For theroectical calculations using software, an estimated trunk section and spur lenghts is acceptable but the final documentation ususally requries the actual lenghts.

I don't agree that if there are no errors on the H1 side, that cable lenghts are ok. You can exceed the maximum cable length for overall or individual spur and still have no communication errors on the H1 side. Even though they might not show up on start up, there would be an increased chance of having an issue show up later if lenghts are exceeded.

To measure each cable via resistance readings would be extremely time comsuming and would require actaully unwiring the segment and would not be feasible. It would be an option but not a practical one.

[jcpilman]

I have never used one, but there are a number of devices which use Time Domain Reflectometry (TDR) to measure cable length. They are not too expensive and should be much more accurate and convenient than a resistance measurement. I had one client who used one to verify the length of every purchased spool of cable.

[type18]

Chris,

I am interested in the details of what you are trying to do.

Are you trying to determine the length of a FF cable that is installed but not connected to anything? It seems like this could be done with a TDR, as previously mentioned.

Or, are you talking about wanting a device that can scan a fully installed FF segment (FF power supply, device couplers, terminators, devices)? Or something in between?

Do you need to document anything else besides length?

Cyrus Kelly
Engineer
Relcom

[CHRISC1024]

It can be either a powered or unpowered segment/section of trunk/spur. I would not want to check the cable lenghts by taking ohm readings as it's not a practical solution. Being able to disconnect trunk sections and shooting each length would be acceptable as this si how it is done for DeviceNet using the Allen Bradley Media Checker. Normally, we would document the lenghts on the segment loop sheet and this would be part of the final document that is stored in the client's system.

I guess we could look at a TDR meter but we have never used one that was not made for the specific application. I will look into what options they give you for defining the cable you are testing. I agree that one should work and is a great idea.

Thanks

[aagostin]

Some questions coming to my mind when reading the suggestions:
* Cable length not relevant when segment is working
Without having spent too much thought on this, I would probably support such statement for the following reasons:
- The cable lengths would not be too far off the design
- Cable lengths are designed with plenty of margins (measure your actual device current and voltage at field device and compare it with the segment calculations!!!)
- Cable lengths are typically not designed to the theoretical limits.
This would conclude to the following:
- Even if the cable lengths would be somewhat longer than planned, it would very likely still be within the theoretical limit
- Even if the cable lengths would be somewhat longer than planned, the segment would have sufficient margin (which can be easily verified by measuring segment current and device voltage)

* TDR:
Would TDR be able to measure a properly terminated segment? According to a brief web research, TDR works based on reflections caused by wire breaks, short, and improperly terminated cables.
And since the terminator is in the device coupler in the field junction box, would there be any signal reflection at all? Would it be possible to see the tiny reflections caused by the spurs?
And what if the spurs have equal length? Would it be possible to identify those?

* Create shorts in the field to measure ohmic resistance
I think this would work, but certainly you require two persons, one in the control room to measure the ohmic resistance and one in the field.
In this case, I would suggest to measure the trunk only.
Spur length impact is negligible (on such things like voltage drop), so if you are not verifying proper wiring of the field device (e.g. by using a FBT-6 right at the device), it does not matter whether the spur is 60m or 80m.


Here are a few additional points to consider:
* Why measure cable length and not verify the actual segment against the design?
I would wonder what do you gain from knowing the actual cable length. The segment is designed in a way that field device total current should be so-and-so, and that resulting from the designed cable length the voltage at the field instrument should be no less than so-and-so.
So if you check the voltage at the field device (for convenience you could simply measure at the junction box; the spur voltage drop is negligible), you would know whether your network operates within the design limits.
What else would you want to know?
And also, what if the actual cable is really longer than as per design? You go back to the calculation tool and re-calculate this segment. So then you may find that with your calculation, the field instrument may only have 9.8 volts instead of 9.85V.
If this is the intention, I would like to ask whether you actually measured the individual field instrument currents and compared those to the designs? Why not? Why only one of the parameters involved in your design? And what about the actual FF power supply voltage? And what about the assumed temperature of the cable?

You are looking at one of many parameters involved in the design of a fieldbus segment, and is not a dominating one. You may have 10% longer cables, but the device current can easily be off by 20% to 30% (if you designed with 20mA plus margins per field instrument). This would support the first suggestion, to ignore the cable length when the segment is working.

[CHRISC1024]

I only ask this as it has been a requirement for the last 2 pharma projects we have been on and was once again a requirement for the project we just bid. Honestly, there is NO way that they are pushing cable limits or even coming close. Documenting spur cable lengths are usually the easy part. Most spurs are premade cables and the part number tells you the length. However, they want to know tteh trunk lenghts. When these are installed on conduit/raceways, it's not practical to walk downn an entire segment trying to follow where the they go and estimate a distance. This should be documented when pulling the cable and be understood it's +/- 10%. For DeviceNeet there is a cable tester but that system has a much shorter allowable length for the trunk and drops.

Most installtion we have doen, the voltage on the trunk and devices has never been below 18V. Noramally a trunk is 500' long and total spur lenght is 150-175'.

If it wasn't required it would not be an issue. I was wondering if anyone ever made this type of device for Fieldbus since it was avaiable for other networks and cables. Also being able to test a cable for shorts would be a great help as well.

[dlancaster]

It is also required to be documented in other industries such as Petrochemical, but in addition, many installation subcontractors get paid by the length of cable installed. It might be a very useful tool for many puropses.

David S. Lancaster, P.E.
Certified Foundation Fieldbus Instructor