Multiply VFD's and Motors

I was told it is not a good idea to run wires for muliply motors controlled by multiple drives in the same conduit. I was looking at using some Altivar drives and going though the instal manuals I could not find anything to back this. Can any one give me some documentation explaining why this is a bad idea? Is it something to do with the potential difference of voltage between two different drives?

Thanks for you help in advance, This will clear up some confusion for me.

AB has a good article Drives-In001g-en-p "wiring and grounding guidelines for pulse width modulation drives (pwm)AC drives.

A drive output is not a sine wave, its a PWM square wave. On a 480V three phase motor on a conventional motor starter the peak phase to phase voltage difference is 554 volts. However, no two phases is at peak at the same time. You can run as many motor leads in the same conduit as can legally fit in the conduit and use 600V rated wires as long as they are all energized from the same source.

But when you have multiple drives now you have a situation where you essentially have a different source for each motor. No longer can you be assured that each is in phase, and that each is running at exactly the same frequency, and you can have a peak to peak voltage ranges on the square wave output (830V IIRC) that exceeds the 600V voltage rating of common wire. You'll cook the insulation and eventually let the smoke out of something (which usually is the expen$ive drive).

If you absolutely are constrained to using the same conduit, then I suggest you look into using drive cable. Drive cable uses XPLE type insulation to withstand the peak voltage differences and it is shielded so that the cables don't cross talk.

Alaric,
Exactly The answer I was looking for. I did some searching but the best I came up with was it was not recomended. Or may cause drive faults due to high frequency switching and increased capacitance.
This left me asking the question Why.....
Each drive I do will have seperate runs of conduit.
Thanks!

jseavers said:

Alaric,
...Or may cause drive faults due to high frequency switching and increased capacitance....
Thanks!

Click to expand...

Yeah, that too. But noise and cross talk is a performance issue, and sometimes you might get away with it. I covered code reasons for not doing it, which is that the peak to peak voltages between different phases of different drives can exceed the insulation ratings of the wires and that there is a way to do it if you use shielded drive cables.

Exceeding voltage rating of wire?

Alaric

I am having a hard time understanding your explanation of why running multiple drive motor leads in the same conduit will exceed the voltage rating of the cable.

On a 480 volt drive, the DC bus voltage will be about 672 (480 x 1.4). Therefore the level of the PWM pulses will be 672 volts. It does not matter if the motor leads in a conduit are from one drive or multiple drives, they will all see the same peak voltage since they are all coming from a DC source with the same voltage. Also remember these are square wave pulses, so each phase has the same voltage level for the total half cycle. Using your reasoning, each motor lead from a single drive should be run in its own conduit because the peak voltage between the wires is higher than 600 volts.

The 600 volt rating of a wire is the RMS voltage, not the peak voltage. The peak voltage of a 480 volt supply is actually 672 volts. Is it ok to use 600 volt wire on a 480 volt line when the peak voltage is actually 672 volts? In Canada, 600 volt line voltages are very common. The peak voltage of a 600 volt supply is 840 volts but it is standard practice to use 600 volt rated wire for these applications including drive motor leads.

Concerning the original question, the following quote is from page 4-12 of the Allen-Bradley document mordred mentioned in the second post.

Within Conduit​

Do not route more than 3 sets of motor leads (3 drives) in the same conduit. Maintain fill rates per applicable electrical codes.

Do not run power or motor cables and control or communications cables in the same conduit. If possible, avoid running incoming power leads and motor leads in the same conduit for long runs.
​

Click to expand...

The main reason for not running multiple drive/motor leads in the same conduit is that the drives will cross-talk due to the high frequency components in the pulses potentially causing false ground, short circuit, and overcurrent faults. Of course, this is somewhat dependent upon how long the joint runs are. You could probably do this for 20 feet and never have a problem but at 100 feet or more, I would expect trouble.

As to the insulation withstand voltage question, I've wondered about this for a long time. 600V rated insulation would expect to see peak voltages of 600 x 1.414 = 848V which is fine for short drive/motor leads. But, due to ringing and reflected wave issues on longer leads, those drive pulses that started out at the drive being 660V high are now potentially over 1200V. That's why MG1P31 motors are tested at 1600V! Clearly, we must routinely be well into the safety factor on the 600V motor leads.

Having said that, I have never experienced 600V drive/motor insulation failure which could remotely be blamed on overvoltage. So, it must be covered somehow.

Maybe someone on here knows how and will tell us.

Isn't that covered by the Carrier frequency? On longer leads all VFD manuals will have their own length per carrier frequency chart. At least thats how I always understood it. The expalnation I had before was that the carrier frequency helps reduce the reflected wave

If you have long runs, it is normally recommended to have output chokes because of ringing and peaks.
Wouldn't the use of output chokes more or less take care of the cross-talk and extreme voltage differences between cables ?

Vic said:

Alaric

I am having a hard time understanding your explanation of why running multiple drive motor leads in the same conduit will exceed the voltage rating of the cable.

On a 480 volt drive, the DC bus voltage will be about 672 (480 x 1.4). Therefore the level of the PWM pulses will be 672 volts. It does not matter if the motor leads in a conduit are from one drive or multiple drives, they will all see the same peak voltage since they are all coming from a DC source with the same voltage. Also remember these are square wave pulses, so each phase has the same voltage level for the total half cycle. Using your reasoning, each motor lead from a single drive should be run in its own conduit because the peak voltage between the wires is higher than 600 volts.

The 600 volt rating of a wire is the RMS voltage, not the peak voltage. The peak voltage of a 480 volt supply is actually 672 volts. Is it ok to use 600 volt wire on a 480 volt line when the peak voltage is actually 672 volts? In Canada, 600 volt line voltages are very common. The peak voltage of a 600 volt supply is 840 volts but it is standard practice to use 600 volt rated wire for these applications including drive motor leads.

Click to expand...

When drive #1 has phase A at max positive peak, and drive #2 has phase A at max negative peak, the voltage differential between the two wires is 1344 V and that exceeds the voltage limitation of THHN wire insulation. This scenario is highly likely in two drives where the leads share the conduit. If you use a drive cable the cable has XPLE type insulation, which is rated for that voltage.

Originally posted by Alaric:

When drive #1 has phase A at max positive peak, and drive #2 has phase A at max negative peak, the voltage differential between the two wires is 1344 V and that exceeds the voltage limitation of THHN wire insulation.

Click to expand...

In order to get a 1344-volt difference you would need a 1344-volt DC bus. The only voltage levels the IGBT output can assume are 672 and 0. So the maximum voltage difference you will see (discounting reflected wave) is 672 volts when one phase is conducting and the other is off.

It's kind of like a high voltage line driver encoder output. The A and A-NOT lines are alternately tied to Vss and COM. A-NOT is not tied to COM and A driven to +Vss and -Vss. The same is done with a PWM output.

Keith

Rms

The peak-to-peak voltage of 480 V. is 1,357.54 V. so insulation rated 600 V. RMS would equal 1,697 V. P-P

John, kamenges is correct. There is no sine wave on the drive output so figuring peak and peak to peak voltages are not relevant. There are only square pulses all equal to 660 volts high. I suppose you could have 660V + to ground and 660V - to ground on adjacent wires but that would be worst case.

The peak-to-peak voltage of 480 V. is 1,357.54 V. so insulation rated 600 V. RMS would equal 1,697 V. P-P

Click to expand...

The RMS line-to-line voltage on a 480 volt supply is 480 volts. The peak voltage is 672 volts. The peak-to-peak line voltage is meaningless as the voltage cannot be at the peak positive and peak negative voltage simultaneously.

When dealing with wire insulation, the voltage to ground is more important than the voltage line-to-line as the only protection is the insulation on one wire. The line-to-line voltage is protected by the insulation on 2 wires. The line to ground voltage is especially critical on ungrounded and impedance grounded systems where the voltage to ground can equal the line-to-line voltage.