hoping someone can provide some insight to this problem.
New citi multi system has been running for 7+ months now - system has two outdoor coils and each has an independent network of indoor units along with a central display and bacnet communications interface to the EMCS/BAS. Recently there have been severe communications errors within the mnet on only one communications loops, not both . Wiring is good (16/2 shld). Factory rep is saying that the issue is because the comm wiring goes from the outdoor coil unit to 4 or so indoor units before the refrigerant distribution box and not from the outdoor coils directly to the refrigerant distribution box then to the indoor units like the mfg drawing shows (don't have the drawing right now else would include part numbers).
I've been involved with a number of citi multi systems in the past and never encountered this problem before - although I personally (as a controls tech) don't like some aspects of their BACNet implementation I have not heard any complaints from end users over the years, and therefore, believe it is a solid product. The mechanical contractor on site has been doing the trouble shooting and relying on the local mfg rep/supplier for support - I think they're grasping at straws.
I have asked my electrical sub-contractor to go back to site and install new cabling as required to make the physical installation match the drawing supplied by mitsubishi. Personally I believe that it will only be a matter of time before the current comm. problem comes back and, if it happedn after the substantial completion date + 364 days, the end user is going to "be on the hook".
Not looking for solution/trouble-shooting tips/tech support but would like some input. If the communications cabling is as per the mfg spec (16/2 shielded), is daisy-chained, and the shield properly grounded at one end only, why would it matter where each device is physically in the comm network? The mfg local rep could not answer this question.