Wiring method for both wet location AND place of assembly

[ckelley3]

I'm a residential guy, but have volunteered to do an HVAC install for my church and need a little help from you commercial guys. Here's the situation.....

Installing 7 Mitsubishi ductless mini splits in the classroom areas of a ca.1930 building. One of the rooms can potentially hold 100 people, so I am planning to follow code for a place of assembly, even though it might be debatable since this is a classroon. All wiring (whether pipe, MC, etc.) will be surface mounted or enclosed in a plastic "line hide" with the refrigerant lines and condensate drain hose. The main challenge is complying with both a wet rated method and a place of assembly method. The wiring between the indoor and outdoor unit is composed of 3 conductors, S1, S2 and S3, plus a ground. S1-S2 is 208/230, with S2-S3 being a 12-24 VDC pulse signal for control. Recently attended their factory install class, in which they are adamant that S1-S2-S3 conductors CANNOT be spliced ANYWHERE between the units. They must be 14AWG, no bigger, no smaller, and can be stranded or solid. They say the communication portion of the circuit can be affected by wire size and splicing, thus their stance. As I see it, I have a few options:

1. Use a PVC coated MC (TECK) cable for a continuous run. Deal with the expense/availability of wet rated connectors outdoors.

2. Use sealtite or NM conduit outside to a JB just inside the interior wall, where the remaining run will be EMT or FMC. Wire with continuous THWN.

3. Use MC for the interior run and NM conduit/sealtite for exterior. Making sure the MC wire is THWN, remove the MC jacket for the portion of wire that needs to be in the conduit/seatite and transition it in a JB just inside the exterior wall.

I seem to be leaning toward option 3. Our church is on a very tight budget, so cost is a major factor. Other ideas/issues?

 

[kwired]

I'm a residential guy, but have volunteered to do an HVAC install for my church and need a little help from you commercial guys. Here's the situation.....

Installing 7 Mitsubishi ductless mini splits in the classroom areas of a ca.1930 building. One of the rooms can potentially hold 100 people, so I am planning to follow code for a place of assembly, even though it might be debatable since this is a classroon. All wiring (whether pipe, MC, etc.) will be surface mounted or enclosed in a plastic "line hide" with the refrigerant lines and condensate drain hose. The main challenge is complying with both a wet rated method and a place of assembly method. The wiring between the indoor and outdoor unit is composed of 3 conductors, S1, S2 and S3, plus a ground. S1-S2 is 208/230, with S2-S3 being a 12-24 VDC pulse signal for control. Recently attended their factory install class, in which they are adamant that S1-S2-S3 conductors CANNOT be spliced ANYWHERE between the units. They must be 14AWG, no bigger, no smaller, and can be stranded or solid. They say the communication portion of the circuit can be affected by wire size and splicing, thus their stance. As I see it, I have a few options:

1. Use a PVC coated MC (TECK) cable for a continuous run. Deal with the expense/availability of wet rated connectors outdoors.

2. Use sealtite or NM conduit outside to a JB just inside the interior wall, where the remaining run will be EMT or FMC. Wire with continuous THWN.

3. Use MC for the interior run and NM conduit/sealtite for exterior. Making sure the MC wire is THWN, remove the MC jacket for the portion of wire that needs to be in the conduit/seatite and transition it in a JB just inside the exterior wall.

I seem to be leaning toward option 3. Our church is on a very tight budget, so cost is a major factor. Other ideas/issues?

I find it very hard to believe you will have the problems they mention if you have splices or changes in conductor size. What do they say about voltage drop if it were an unusually long run?

That said if you want to run continuous conductors and want to keep cost down, I would probably stick with commonly available methods, or specialty items will cost you even if they are easier to install.

PVC coated MC may be expensive - especially if you have to buy a minimum quantity and will not use the majority of it, plus it likely won't fly inside the place of assembly. If you remove the sheath for the interior portion and sleeve with an acceptable raceway, the conductors must be a type recognized by NEC as well as marked. Some conductors in MC cable are marked but not all are.

 

[iwire]

I would splice without a moments hesitation

 

[sparkyrick]

If the installation instructions say "no splices" and you splice, you may be voiding the warranty.

Just sayin'

 

[iwire]

If the installation instructions say "no splices" and you splice, you may be voiding the warranty.

Just sayin'

I would splice, just sayin'

 

[Volta]

Why not option #4 if you really don't want to splice?

EMT.

 

[K8MHZ]

I would splice, just sayin'

And totally ignore this?

They say the communication portion of the circuit can be affected by wire size and splicing,

 

[iwire]

And totally ignore this?

Yes.

 

[K8MHZ]

Yes.

Is that to say you are certain that splices and wire size will not, under any circumstances, adversely affect communication and the factory instructors are wrong?

Or is communication quality just of no concern to you?

 

[ckelley3]

I find it very hard to believe you will have the problems they mention if you have splices or changes in conductor size. What do they say about voltage drop if it were an unusually long run?

That said if you want to run continuous conductors and want to keep cost down, I would probably stick with commonly available methods, or specialty items will cost you even if they are easier to install.

PVC coated MC may be expensive - especially if you have to buy a minimum quantity and will not use the majority of it, plus it likely won't fly inside the place of assembly. If you remove the sheath for the interior portion and sleeve with an acceptable raceway, the conductors must be a type recognized by NEC as well as marked. Some conductors in MC cable are marked but not all are.

The FLA of the indoor units (1.6A) makes the potential for voltage drops pretty much non-existent. Also, there is a maximum line set length for the refrigeration circuit that is limiting.

They probably went over the whole wire size and splicing thing 5--6 times in the training, even stating they've seen systems act up with #12 wire instead of #14 for this circuit, so I am taking this requirement seriously.

 

[augie47]

The whole mini-split scene has been a nightmare for many inspectors and electricians.
Many of the systems are foreign and many are covered by various listing agencies. Some come with instructions calling for wiring methods that are not NEC allowed (cord), most ignore the need for a disconnecting means on the interior motor and some of the larger units ignore the fact that the smaller conductors to the interior unit do not, on some units, have overcurrent protection.
Add to that a myriad of "special" instructions added by the manufacturer.
I talked with some UL folks concerning some of the problems where manufacture instructions clashed with NEC and they noted the units were not UL listed but rather ETL. ETL folks said they would "look into it" (twice).
FWIW, I know of several Mitsubishi splits that have had switches added and wires spliced that seem to be operating fine... (not saying your unite would)

 

[George Stolz]

Is that to say you are certain that splices and wire size will not, under any circumstances, adversely affect communication and the factory instructors are wrong?

Or is communication quality just of no concern to you?

Not to speak for Bob, but I've noticed a real problem of people not completing the simple task of installing a wire nut without a problem. This would result in service calls where a splice was at fault, which will result in the factory issuing draconian proclamations like "no splice allowed."

If you're one of the dying breed who still knows how to install a wire nut, then I would not have a moments hesitation doing so. The physics are simple, there is no way for the equipment to know you've made a splice if you've done your job properly.

 

[kwired]

Is that to say you are certain that splices and wire size will not, under any circumstances, adversely affect communication and the factory instructors are wrong?

Or is communication quality just of no concern to you?

As I see it you could run an unspliced line and still have a poor connection at one end or the other cause the same problem. The HVAC guys are just trying to make less troubleshooting when something goes wrong. I don't know how many times I have had some HVAC guy blame a raceway type equipment grounding conductor to cause problems with their flame sensor. The explanation of why they need a copper EGC all the way back to the panel makes sense, but almost every time I come in and add a copper EGC the problems they were having did not go away, now I am not so stupid anymore when it is done the way they want it and it still doesn't work:slaphead:

The FLA of the indoor units (1.6A) makes the potential for voltage drops pretty much non-existent. Also, there is a maximum line set length for the refrigeration circuit that is limiting.

They probably went over the whole wire size and splicing thing 5--6 times in the training, even stating they've seen systems act up with #12 wire instead of #14 for this circuit, so I am taking this requirement seriously.

Most of what I said above applies to this also. I am not a refrigeration expert, but I am willing to bet the size of the lines can be increased for longer runs and it will work fine, it also will need more refrigerant to fill the extra space in the system, this is somewhat simple hydraulics/pneumatics application here.

The whole mini-split scene has been a nightmare for many inspectors and electricians.
Many of the systems are foreign and many are covered by various listing agencies. Some come with instructions calling for wiring methods that are not NEC allowed (cord), most ignore the need for a disconnecting means on the interior motor and some of the larger units ignore the fact that the smaller conductors to the interior unit do not, on some units, have overcurrent protection.
Add to that a myriad of "special" instructions added by the manufacturer.
I talked with some UL folks concerning some of the problems where manufacture instructions clashed with NEC and they noted the units were not UL listed but rather ETL. ETL folks said they would "look into it" (twice).
FWIW, I know of several Mitsubishi splits that have had switches added and wires spliced that seem to be operating fine... (not saying your unite would)

That is what I have experienced for the most part.

 

[K8MHZ]

If you're one of the dying breed who still knows how to install a wire nut, then I would not have a moments hesitation doing so. The physics are simple, there is no way for the equipment to know you've made a splice if you've done your job properly.

The OP mentioned that the conductors were for communication equipment. I can definitely see how splices could affect a communication system that relied on precise timing. They do add resistance (very little, but in a precise system that may not be acceptable) and that resistance will not remain a constant. If the data is being transmitted at a rapid pace (much faster than 60 Hz) reactance at splices may become an issue which isn't at 60 Hz.

What works perfect for power and control systems may not be adequate for communications systems.

 

[kwired]

The OP mentioned that the conductors were for communication equipment. I can definitely see how splices could affect a communication system that relied on precise timing. The do add resistance (very little, but in a precise system that may not be acceptable) and that resistance will not remain a constant.

What works perfect for power and control systems may not be adequate for communications systems.

My guess is the signal in the op is more of a control signal than a data signal, splicing the conductor should not have much impact on operation, after all it already has a termination of some kind at each end of it anyway. If the signal was that sensitive the instructors in the OP's situation would have spent time addressing just how to make those terminations.

 

[George Stolz]

The OP mentioned that the conductors were for communication equipment. I can definitely see how splices could affect a communication system that relied on precise timing.

We're talking a split system AC, not the space station.

 

[K8MHZ]

We're talking a split system AC, not the space station.

But that type of complexity is working it's way into just about everything made today.

OP mentions a pulse signal for control. Also mentioned is that there is 12-24 VDC from one line to another, that which is one leg of an AC voltage 10 times higher. If pulse refers to pulse width modulation, that is basically FM. Perhaps their form of PWM relies on a consistent velocity factor in order to function. A splice will definitely affect velocity factor.

If I were a student in that class, after being told several time not to splice, I would have wanted a detailed explanation as to why not. I do know that after being told repeatedly by representatives from the factory not to splice, I would heed their words, not ignore them based upon something I read on an Internet forum.

 

[iwire]

Is that to say you are certain that splices and wire size will not, under any circumstances, adversely affect communication and the factory instructors are wrong?

Or is communication quality just of no concern to you?

Having wired many Mitsubishi split systems I can say with confidence they work fine with splices.

 

[iwire]

My guess is the signal in the op is more of a control signal than a data signal, splicing the conductor should not have much impact on operation, after all it already has a termination of some kind at each end of it anyway. If the signal was that sensitive the instructors in the OP's situation would have spent time addressing just how to make those terminations.

They are becoming data signals more than control signals.

The outdoor unit has to recognize the indoor unit and they 'talk' to each other.

 

[iwire]

But that type of complexity is working it's way into just about everything made today.

Which is true, but if the level of complexity is getting so great that feild splicing cannot be done that is a very bad design.

Of course I am not a fan of these mini splits anyway for the reasons Gus mentioned. They are really not designed for the American market and codes.