Wire insulation code question; VFD/Chillers and RS485 = 600v??

[maxburn]

Location is North Carolina and we have some basic RS485 BACnet MS/TP communication going to VFD/Chiller/RTU etc etc. Inspector sees this equipment is wired for 480v and says every wire in the system needs to be 600volt rated. Problem is the RS485 cable sold everywhere is only 300v rated. Seems some of the 600v rated comm wire exists but it's like 8x cost increase and hardly available anywhere. Outside looking in I'm thinking to my self if everyone is actually required to use this then the cable would exist in reasonable price and availability. I suspect we are being messed with but I don't know NEC.

Question; what exceptions would NEC carry to allow comm wire at 300v insulation rating in a VFD or Chiller? Looking at these VFD's they don't really have a raceway or partition between the comm entrance and the motor/line wiring entrance.

 

[tom baker]

For the Ethernet cooks on the drives I wired, ran a separate conduit for Ethernet. cable into the drive was kept away from power conductors. Art 725 allows class 2 wiring in same enclosure if separation is maintained.

 

[petersonra]

run the comm cables in smurf tubing. no question than about it being separate.

 

[tom baker]

run the comm cables in smurf tubing. no question than about it being separate.

I have done similar with Carflex thru bottom of drive from raceway below.

 

[maxburn]

Ethernet sounds like what we are doing. We do have a individual conduit/flex for the comm wiring and a couple other things like start/stop and analog 0-10v signal. Piecing that together I believe we are "Class 1 Power-Limited Circuits".

VFD's in particular don't have a cable pathway that separates the low voltage from the high voltage, which is the problem. I know it's a bit of work but is there something specific in 725 I can cite?

I don't know what smurf tubing is, please explain.

I do know that the inspector already shot down 600v rated heat shrink citing something that cables can't be upgraded.

 

[Dennis Alwon]

Ethernet sounds like what we are doing. We do have a individual conduit/flex for the comm wiring and a couple other things like start/stop and analog 0-10v signal. Piecing that together I believe we are "Class 1 Power-Limited Circuits".

VFD's in particular don't have a cable pathway that separates the low voltage from the high voltage, which is the problem. I know it's a bit of work but is there something specific in 725 I can cite?

I don't know what smurf tubing is, please explain.

I do know that the inspector already shot down 600v rated heat shrink citing something that cables can't be upgraded.

This is smurf tubing
aka ent-- electrical non metallic tubing

 

[maxburn]

OK. Our electrician has already mentioned that these VFD enclosures do not have enough room to run flex inside them up to the low voltage section.

 

[petersonra]

OK. Our electrician has already mentioned that these VFD enclosures do not have enough room to run flex inside them up to the low voltage section.

If there is room for the Ethernet cables there is room for the Smurf tubing or something similar

 

[maxburn]

I hear you, there's a couple things in this story that doesn't line up. Just what I'm dealing with here. I could see they can get flex in the bottom past the high voltage but not all the way up to the low voltage terminal strip. You'd think once past the high voltage section the inches of separation clause would kick in...

I did call Belden and they recommended 3074F. Windy City Wire didn't have a equivalent for that but they are looking. Part of the issue is our vendor specifically wants 22ga comm wire and that 600v stuff is going to be 18ga. BUT RS485 spec and BACnet 9.2 seem fine with 18ga as long as a list of other things are met. Personally I think 18ga is going to make long runs work better for BIAS reasons. This doc explains that; https://www.ccontrols.com/pdf/Extv9n2.pdf

 

[petersonra]

725.136 Separation from Electric Light, Power, Class 1, Non–
Power-Limited Fire Alarm Circuit Conductors, and Medium-
Power Network-Powered Broadband Communications Cables.
...
(B) Separated by Barriers. Class 2 and Class 3 circuits shall be
permitted to be installed together with the conductors of electric
light, power, Class 1, non–power-limited fire alarm and
medium power network-powered broadband communications
circuits where they are separated by a barrier.

Plastic tubing of some sort is a barrier.

 

[maxburn]

Yeah, define barrier.

This inspector already turned down 600v rated shrink tubing. Cited Code reference article 300 where insulation must match voltage

we pitched that because it actually carried that 600v rating rather than random plastic tubing and flex that would be harder to install anyway.

 

[petersonra]

Yeah, define barrier.

This inspector already turned down 600v rated shrink tubing. Cited Code reference article 300 where insulation must match voltage

we pitched that because it actually carried that 600v rating rather than random plastic tubing and flex that would be harder to install anyway.

I am inclined to agree that heat shrink is not an answer, but his reasoning is just plain wrong. It increases the insulation to 600 V but article 725 does not allow one to mix class 2 and power circuits just because the voltage rating of the insulation of the class 2 conductors is adequate.

If what he is saying is true you could not use heat shrink or black tape to repair insulation that has been damaged.

 

[maxburn]

I didn't like the heat shrink idea much either but it had the magic number on it and would have been easy.

It feels like there are some grounds somewhere for the 600v insulation makes the difference, maybe NFPA 70 and talking about regular ethernet here; https://www.belden.com/blogs/600v-industrial-ethernet-cable-safety-and-nec-considerations

I would start looking into what defines "barrier" but I'm told we ordered the super expensive special cable this morning so it would be too late. There's probably an argument to be made there but we'd probably just catch problems elsewhere.

 

[petersonra]

I didn't like the heat shrink idea much either but it had the magic number on it and would have been easy.

It feels like there are some grounds somewhere for the 600v insulation makes the difference, maybe NFPA 70 and talking about regular ethernet here; https://www.belden.com/blogs/600v-industrial-ethernet-cable-safety-and-nec-considerations

I would start looking into what defines "barrier" but I'm told we ordered the super expensive special cable this morning so it would be too late. There's probably an argument to be made there but we'd probably just catch problems elsewhere.

IIRC, the reason you can use the expensive ethernet cable is because UL has defined the outer shell as a barrier, and not just insulation.

 

[SceneryDriver]

Location is North Carolina and we have some basic RS485 BACnet MS/TP communication going to VFD/Chiller/RTU etc etc. Inspector sees this equipment is wired for 480v and says every wire in the system needs to be 600volt rated. Problem is the RS485 cable sold everywhere is only 300v rated. Seems some of the 600v rated comm wire exists but it's like 8x cost increase and hardly available anywhere. Outside looking in I'm thinking to my self if everyone is actually required to use this then the cable would exist in reasonable price and availability. I suspect we are being messed with but I don't know NEC.

Question; what exceptions would NEC carry to allow comm wire at 300v insulation rating in a VFD or Chiller? Looking at these VFD's they don't really have a raceway or partition between the comm entrance and the motor/line wiring entrance.

RS485 works just fine over Ethernet cable. Use the fancy (expensive) 600V shielded CAT6 cable. One twisted pair for D+ and D-, and use another pair along with the shield as the common.

We run DMX - which on the physcal later is RS485 at 250Kbaud - over CAT cable all the time and it works perfectly. The characteristic impedance of CAT cable is 110ohm, and RS485 usually wants 100ohm. So little diffreence it won't matter.

CAT6 is 23awg wire, so slightly smaller than the 22awg your vendor wants but the difference is so small it shouldn't matter. Just use ferrules on the wires when you terminate in the terminal blocks.


SceneryDriver

 

[paulengr]

Paul from EMS in Wake Forest.

That super overpriced Belden cable is Listed as 600 V AWM. AWM or appliance wiring media is a generic “pass through” UL Component Listing. It is the same Listing as you will see on the motor lead wiring. So we have a situation where UL says everything you use to make something is either raw materials, common hardware, a “component”, or an “assembly”. Assemblies are final products. That’s the stuff electricians can use as long as it complies with NEC. “Components” can’t be used as final products…they must be used as part of an assembly. UL stamps the component as an intermediate product because it would be hard to test it as a final assembly or because it’s more than just a raw material,

So AWM is component wire. It can be used as say the raw SJOOW that is the cable used to make molded extension cords, motor or transformer lead wires, lamp wiring, etc. NONE of it meets NEC and isn’t a legal type by Code. So say Frick can run AWM 600 V Ethernet cable inside the compressor anywhere and it’s legal. But you can’t use it except replacement in kind.

Now category cable is legal as communication cable and falls under chapter 7. It can be run in power limited cable trays and conduits. But because it is not #18 or larger it cannot be run in general purpose trays and conduits under chapter 3, period. Even though it is legal tie wrapped to power cables inside the control panel it can’t leave the panel like that. If all you want is Modbus just use ITC or fire alarm cable. There are tons of 2-4 conductor foil wrapped options that are #18.

 

[SAAPI]

RS485 works just fine over Ethernet cable. Use the fancy (expensive) 600V shielded CAT6 cable. One twisted pair for D+ and D-, and use another pair along with the shield as the common.

We run DMX - which on the physcal later is RS485 at 250Kbaud - over CAT cable all the time and it works perfectly. The characteristic impedance of CAT cable is 110ohm, and RS485 usually wants 100ohm. So little diffreence it won't matter.

CAT6 is 23awg wire, so slightly smaller than the 22awg your vendor wants but the difference is so small it shouldn't matter. Just use ferrules on the wires when you terminate in the terminal blocks.


SceneryDriver

Oriental motor stepper motor drivers put a terminal strip in parallel with a Cat5 connector - it is so sweet to just use CAT 6 wires for hook up.

 

[tortuga]

I imagine, just as with anything AHJ's dont see, catch, or enforce
if there is an arc blast, or a fire or meltdown or whatever in a 480V enclosure and some $250,000 piece of equipment is damaged,
the customer/business owners insurance forensics investigator would catch it and pursue subrogation recovery against against the insurance of the contractor that installed it.

 

[Jraef]

Rockwell has an Ethernet cable that is UL listed as PLTC at 600V, meaning it can be run separately by itself in a panel, raceway or tray (gets around that AWM issue).

 

[oceanobob]

If the 480 circuit is comprised of perhaps two or maybe three ungrounded conductors - each conductor is less than 300 volts to reference. Maybe about 277. For comparison: a 575 volt circuit would be about 350 or more to reference.

The article 300 for conductors of different systems indicates insulation value is max circuit voltage to any conductor. Because this is singular, and one conductor (of a typical 480 circuit) cannot itself produce 480 volts - perhaps the voltage value attribute is for any (one) conductor compared to reference.

With this perusal.... a 480 circuit, when placed together with say ethernet cable with 300 volt insulation rating, would meet the criteria for the ethernet insulation to be 300 volts - but may not be ok for the 575 volt system.


This particular line of reasoning was explained as to why the 480 volt VFD with all the listings etc has the connections for the various circuits with no particular regard to separation etc.