Page 2 of 2 FirstFirst 12
Results 21 to 35 of 35

Thread: BACnet MS/TP + Bad Wire, will it work?

  1. #21
    Join Date
    Feb 2005
    Location
    Edmonton, AB Canada
    Posts
    876
    Post Likes
    Quote Originally Posted by jschulze View Post
    There is quite a lot of information on RS-485 available online. I've compiled a pretty large list of links, but here are some of the most useful. The first link is a basic FAQ on RS-485 and the content provided in the links get more advanced as the list goes down.

    https://www.csimn.com/CSI_pages/RS-485-FAQ.html

    https://www.ti.com/lit/an/slla272c/slla272c.pdf

    https://www2.htw-dresden.de/~huhle/ArtScienceRS485.pdf

    https://www.ti.com/lit/an/slla070d/slla070d.pdf


    Additional app note on RS-485 grounding:
    http://robustdc.com/download/index.php?file=AN005.pdf

    Additional article on Termination and Biasing
    https://blog.opto22.com/optoblog/rs-...e-bias-or-both

    Good day jschulze,

    Indeed a solid understanding of RS485 is very important. It is unfortunate that that the info you presented appears to not have been read or understood by a number of the equipment manufacturers...

    Cheers,

    Sam

  2. Likes bigguy158 liked this post.
  3. #22
    Join Date
    Oct 2018
    Location
    WI
    Posts
    77
    Post Likes
    Quote Originally Posted by s2sam View Post
    Good day jschulze,

    Indeed a solid understanding of RS485 is very important. It is unfortunate that that the info you presented appears to not have been read or understood by a number of the equipment manufacturers...

    Cheers,

    Sam
    Yes, I would agree. Many manufacturers of RS-485 equipment, and even some techs that commission RS-485 networks, do indeed seem to lack an understanding of RS-485.

    I can't believe the number of RS-485 devices I've seen that only have + and - terminals with no mention of what those terminals are referenced to, whether it's isolated, etc. If they do have an additional terminal, many times it's labeled as shield (sometimes incorrectly, as the terminal is often connected to the device's logic ground). There seems to be an awful lot of manufactures that don't understand that there is a big difference between shield and common-mode (0V) reference.

    I've also worked with techs from one of the big-name control companies that stated they do have a reference wire on their network, but they don't land it on any other manufacturers' devices because the company has a policy that forbids them from landing the reference wire onto 3rd party devices.

    It's no wonder why there's so many issues in the field with RS-485 interoperability between manufacturers...

  4. #23
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    3,081
    Post Likes
    Quote Originally Posted by jschulze View Post
    I've also worked with techs from one of the big-name control companies that stated they do have a reference wire on their network, but they don't land it on any other manufacturers' devices because the company has a policy that forbids them from landing the reference wire onto 3rd party devices.
    Sounds like Johnson - good job being discreet, but I am gonna call that.
    I remember doing internal Johnson training for Bacnet. The intense impression I got was that there is Johnson Bacnet and everything else, and everything else is a horrifying morass of inconsistent unpredictability and anything having to do with it is a roll of the dice at best....
    So I wuda left my 3rd wire float too with that understanding.
    Hmmmm....smells like numbatwo to me.

  5. #24
    Join Date
    Feb 2005
    Location
    Edmonton, AB Canada
    Posts
    876
    Post Likes
    Quote Originally Posted by jschulze View Post
    Yes, I would agree. Many manufacturers of RS-485 equipment, and even some techs that commission RS-485 networks, do indeed seem to lack an understanding of RS-485.

    I can't believe the number of RS-485 devices I've seen that only have + and - terminals with no mention of what those terminals are referenced to, whether it's isolated, etc. If they do have an additional terminal, many times it's labeled as shield (sometimes incorrectly, as the terminal is often connected to the device's logic ground). There seems to be an awful lot of manufactures that don't understand that there is a big difference between shield and common-mode (0V) reference.

    I've also worked with techs from one of the big-name control companies that stated they do have a reference wire on their network, but they don't land it on any other manufacturers' devices because the company has a policy that forbids them from landing the reference wire onto 3rd party devices.

    It's no wonder why there's so many issues in the field with RS-485 interoperability between manufacturers...
    Good day jschulze,

    Well said and I agree.

    Cheers,

    Sam

  6. #25
    Join Date
    Feb 2005
    Location
    Edmonton, AB Canada
    Posts
    876
    Post Likes
    Quote Originally Posted by numbawunfela View Post
    Sounds like Johnson - good job being discreet, but I am gonna call that.
    I remember doing internal Johnson training for Bacnet. The intense impression I got was that there is Johnson Bacnet and everything else, and everything else is a horrifying morass of inconsistent unpredictability and anything having to do with it is a roll of the dice at best....
    So I wuda left my 3rd wire float too with that understanding.

    Good day numbawunfela,

    I cannot speak for jschulze or any inferences he may have made, but I can tell you that JCI's interfaces are one of the better designed ones I have seen. I have not analyzed (or dismembered ) all of JCI's devices, but for the past 20+ years I have looked at a lot of their devices and found that they were one of a very few number of manufacturers that electrically isolated their RS485 (N2 Bus) interfaces... and provided pretty beefy I/O protection circuitry. This is even more impressive considering that at the time implementing electrically isolated RS485 interfaces was not as straightforward as it is today. Electrically isolating the RS485 has significant benefits and can reduce communication issues between other manufacturer devices... and so the JCI Field device designers were certainly thinking. Going forward, I still periodically "review" JCI's offerings and although their software/firmware may be wanting, their hardware seemed to maintain the same standard they had. Hopefully this will continue, but time will tell and also will indicate if bean counters or unseasoned designers are driving their field devices.

    Cheers,

    Sam

  7. #26
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    3,081
    Post Likes
    Quote Originally Posted by s2sam View Post
    I cannot speak for jschulze or any inferences he may have made, but I can tell you that JCI's interfaces are one of the better designed ones I have seen.
    I am sure I have not gotten into the level of detail you have, but I tend to agree. Johnson hardware is pretty good.
    I was more commenting on the internal culture that may make it seem reasonable for one of their guys to think it was a good idea to avoid having one of their devices become 'tainted' by another vendor's bacnet device - and avoid landing that 3rd wire anywhere.
    Hmmmm....smells like numbatwo to me.

  8. #27
    Join Date
    Feb 2005
    Location
    Edmonton, AB Canada
    Posts
    876
    Post Likes
    Quote Originally Posted by numbawunfela View Post
    I am sure I have not gotten into the level of detail you have, but I tend to agree. Johnson hardware is pretty good.
    I was more commenting on the internal culture that may make it seem reasonable for one of their guys to think it was a good idea to avoid having one of their devices become 'tainted' by another vendor's bacnet device - and avoid landing that 3rd wire anywhere.
    Good day numbawunfela,

    Gotcha... tells you where my brain is at... hardware/software/firmware. Corporate culture and politics are not part of my makeup and so explains why I never found a corporate home.

    Cheers,

    Sam

  9. #28
    Join Date
    Sep 2013
    Posts
    695
    Post Likes
    This is a interesting discussion, hopefully you find this question interesting. I currently am dealing with a comm issue at a hospital we service.

    The trunk is two JCI bacnet devices and 8 Setra SRPM room pressure monitors, less than 1000’ long. Wire is 3 conductor shielded. We added a domestic water pump skid to the trunk and some devices immediately went offline.

    We ended up repulling part of the trunk to remove a splice and put one of the JCI devices at the end of the line so we could turn on its EOL switch. This made things better and we were able to map in the pump skid. I also added a MS-BACEOL at the Jace.

    The customer contacted me after 2 months and said things would still go offline occasionally. Before adding the pump skid the trunk was solid.

    I’m going to go there today and get the pump skid info. The skid is only two wire and I did not see a spot to land the reference. It sounds like it could be the reference for this pump skid is using the earth ground and not working so hot?

    My question is there something I can do to land the reference on this device if it is only two wire?


    Sent from my iPad using Tapatalk

  10. #29
    Join Date
    Sep 2013
    Posts
    695
    Post Likes
    Quote Originally Posted by s2sam View Post
    Good day numbawunfela,

    Gotcha... tells you where my brain is at... hardware/software/firmware. Corporate culture and politics are not part of my makeup and so explains why I never found a corporate home.

    Cheers,

    Sam
    So questions, if the device is non isolated and the reference wire is hooked up won’t that cause ground loops?

    Therefore if it is really hard to know if a device is non isolated or only partially isolated maybe corporates knee jerk response was to never land the third wire?


    Sent from my iPad using Tapatalk

  11. #30
    Join Date
    Sep 2013
    Posts
    695
    Post Likes
    Quote Originally Posted by jschulze View Post
    Yes, I would agree. Many manufacturers of RS-485 equipment, and even some techs that commission RS-485 networks, do indeed seem to lack an understanding of RS-485.

    I can't believe the number of RS-485 devices I've seen that only have + and - terminals with no mention of what those terminals are referenced to, whether it's isolated, etc. If they do have an additional terminal, many times it's labeled as shield (sometimes incorrectly, as the terminal is often connected to the device's logic ground). There seems to be an awful lot of manufactures that don't understand that there is a big difference between shield and common-mode (0V) reference.

    I've also worked with techs from one of the big-name control companies that stated they do have a reference wire on their network, but they don't land it on any other manufacturers' devices because the company has a policy that forbids them from landing the reference wire onto 3rd party devices.

    It's no wonder why there's so many issues in the field with RS-485 interoperability between manufacturers...
    How easy is it find out how the device BaCnet reference e is handled?

    Here is a proposal: if you want to be BTL listed you must declare in your BIBB or PIC what you are using for your RS485 transceiver chip?


    Sent from my iPad using Tapatalk

  12. #31
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    3,081
    Post Likes
    Bigguy, 2 wire devices talk on RS485 - a 3 wire medium - using ground as their 3rd conductor. Johnson and Siemens use their 3rd wire, and isolated transceivers. Isolated as in - they are not looking at a grounded secondary on the power to provide that 3rd conductor, the transceiver is isolated from the power and therefore has his own 3rd conductor.
    When mixing the 2, you can land the 3rd wire at a non-isolated device by grounding it through a resistor. I have seen documentation that says 100ohms. I most often use a 120 ohm resistor as that is what I have laying around for termination. Then the non-isolated device will share a 3rd wire with the isolated section. To avoid ground loops I land it at one place, usually the JACE. Then you are as good as any other trunk.... you are relying on the ground of the building to get the 2 wire stuff.
    I also will use 2 of the MS-BASCEOL devices, one at each end on problem trunks. If you only have one, set the JACE to BIA and add a termination resistor. Then put the MS-BASCEOL at the other end.
    A voltmeter or Oscilloscope will tell you if these electrical properties are the issue.but there is a whole other side to this, and that is the Wireshark side.
    Hmmmm....smells like numbatwo to me.

  13. Likes jschulze liked this post.
  14. #32
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    3,081
    Post Likes
    Guys, I have 3 - one hour classes I did for my shop on doing this. I cover everything up to (but not including) the Oscilloscope and Wireshark parts. I did it for my internal shop training. Would there be an interest in a pay-per-view model that is put at a reasonable level? I have been taking my time in getting it put up, but I can see this is hell for a lot of people.
    I have video of real sites with the voltmeter and showing how I fix these exact issues.
    I have all the pieces for an Oscilloscope class, but have not assembled it.
    It will not be available before you get onsite today, Bigguy.... sorry.
    Just trying to guage interest.
    Hmmmm....smells like numbatwo to me.

  15. #33
    Join Date
    Oct 2018
    Location
    WI
    Posts
    77
    Post Likes
    Quote Originally Posted by bigguy158 View Post
    So questions, if the device is non isolated and the reference wire is hooked up won’t that cause ground loops?

    Therefore if it is really hard to know if a device is non isolated or only partially isolated maybe corporates knee jerk response was to never land the third wire?


    Sent from my iPad using Tapatalk
    You're exactly right, which is why there's not just one solution for how to ground all RS-485 networks. It depends on the devices being used, the location's (building's) grounding network, cabling, and so many more things.

    Have a look at the articles I posted in post #18. Almost all of them cover RS-485 grounding techniques.

  16. #34
    Join Date
    Oct 2018
    Location
    WI
    Posts
    77
    Post Likes
    Quote Originally Posted by bigguy158 View Post
    How easy is it find out how the device BaCnet reference e is handled?

    Here is a proposal: if you want to be BTL listed you must declare in your BIBB or PIC what you are using for your RS485 transceiver chip?


    Sent from my iPad using Tapatalk
    I like the idea, but there are a lot of other protocols that use RS-485 besides BACnet too. Also, communication protocol specification and conformance tests don't typically cover the physical RS-485 layer (it's outside the scope of the protocol). There is already a standard for that: EIA/TIA-485.

  17. Likes VAEngineer liked this post.
  18. #35
    Join Date
    Oct 2018
    Location
    WI
    Posts
    77
    Post Likes
    The good news in all of this is that RS-485 transceiver technology is advancing and many manufacturers are now isolating their RS-485 circuitry. Isolation solves a lot of the RS-485 grounding problems, including ground loops, and will most often even work without the reference wire connected at all.

    Many modern RS-485 transceivers provide full-fail safe operation, keeping their inputs at a known state even when the the bus is open or shorted. This removes the need for network biasing and prevents issues caused by applying termination to an unbiased bus (i.e. the + and - idle differential voltage is below 200mV).

    Additionally, most modern transceivers consume only a fraction of a unit load (1/2, 1/4, 1/8, etc.) allowing for many more than 32 devices on the bus (64, 128, 256, etc.) without using repeaters.

    There are also additional features available in some transceivers, such as high output drive (allowing longer cable runs and more than 32 unit loads) and slew rate limiting (which limits the rise time of transmitted transitions, preventing cable reflections, and thus removing the need for termination resistors - although termination may still be necessary for very long cable runs and high baud rate combinations).

  19. Likes JCIman, VAEngineer liked this post.
Page 2 of 2 FirstFirst 12

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •