Wave Phenomena and VFD's

[DMG_1]

I have been doing some research on VFD's and I have come across the subject of wave phenomena due to long cable lengths. I have found equations on determining voltage spikes seen at motor terminals based on characteristic impedance but I am having problems finding cable data for capacitance to use in the equation. I believe I can calculate cable inductance based on inductive reactance as shown in NEC Chapter 9 Table 9. I have not been able to find values of capacitance, anyone out there know where I can find this information?

 

[GoldDigger]

I have been doing some research on VFD's and I have come across the subject of wave phenomena due to long cable lengths. I have found equations on determining voltage spikes seen at motor terminals based on characteristic impedance but I am having problems finding cable data for capacitance to use in the equation. I believe I can calculate cable inductance based on inductive reactance as shown in NEC Chapter 9 Table 9. I have not been able to find values of capacitance, anyone out there know where I can find this information?

If you cannot find tables, it is possible to calvulate it reasonably well from the conductor radius, the insulation dielectric constant, and the cable geometry.
Or you could use an impedance bridge and a short section of cable. Or you could just measure the transmission line impedance directly by terminating it with an various resistances and using a TDR.

 

[gar]

131001-2400 EDT

DMG_1:

Can you define what is meant by "wave phenomena"?

If you are talking about waveforms that result from transmission lines that are not terminated in the characteristic impedance of the line, then these result from reflected energy from the improperly terminated line and frequencies resulting from a step change will depend upon the line length and propogation velocity of energy along the line (generally around about 0.7 times the speed of light). See some waveforms at my site http://beta-a2.com/cat-5e_photo.html .

150 ft puts you in the 2 to 5 MHz range, and 1000 ft around 300 to 800 kHz.

Otherwise you probably treat the circuit as lumped capacitance and inductance. This is probably not the transmission line inductance.

.

 

[GoldDigger]

131001-2400 EDT

DMG_1:

Can you define what is meant by "wave phenomena"?

If you are talking about waveforms that result from transmission lines that are not terminated in the characteristic impedance of the line, then these result from reflected energy from the improperly terminated line and frequencies resulting from a step change will depend upon the line length and propogation velocity of energy along the line (generally around about 0.7 times the speed of light). See some waveforms at my site http://beta-a2.com/cat-5e_photo.html .

150 ft puts you in the 2 to 5 MHz range, and 1000 ft around 300 to 800 kHz.

Otherwise you probably treat the circuit as lumped capacitance and inductance. This is probably not the transmission line inductance.

.

What the OP is talking about is reflections (bad) and standing waves (worse) that typically happen on wiring between a load and a VFD or other voltage source that has fast rise time waveforms. If you know the characteristic impedance of the wiring considered as multiple transmission lines and terminated by a known partially reactive impedance (the motor plus any PFCC), you can approximate the nodal voltage and current maxima knowing only that the driving waveform will contain high enough harmonics to approximate the standing wave frequencies.
These component frequencies of the fast edges are high enough that you can sometimes observe bands of overheated or voltage-stressed insulation at intervals far shorter than the total length of the wire. You are not, of course, looking for resonances at 60 Hz or low harmonics.

Note also that if you know the lumped inductance and capacitance per foot, you also know the distributed inductance and capacitance and you can calculate the characteristic resistance of the transmission line from that.

Well shielded cable with extra insulation voltage capability is one way of dealing with this and lumped component harmonic filters (low pass) are another.

PS: To minimize power losses (heating) in the switching elements of the VFD or other supply at high current and high DC input voltage, the switching times must be kept very very short. Comparable to or higher dv/dt than you see in an RS-232 driver circuit.

 

[DMG_1]

I am talking about the characteristic impedance of the cabling running from the VFD to the motor. A reflected wave is seen at the motor terminals due to an impedance mismatch between the cabling from the VFD to the motor. The link to one of the documents I have referenced for wave phenomena is below.

http://www.geindustrial.com/publibrary/checkout/FETP101?TNR=Application and Technical|FETP101|PDF

Equation (1) on page 1 and Equation (3) on page 3 require values for L and C. I would assume that I can calculate L from NEC Chapter 9 Table 9 using XL. However, I have not been able to find any tables or data indicating conductor capacitance or Xc.

 

[gar]

131002-0941 EDT

DMG_1:

The characteristic impedance is probably between 50 and 300 ohms. Do you really care other than for a theoretical analysis to demonstrate why there are transient peaks. What you want is a way to remove or dampen the spikes. In my waveforms I showed some of the characteristics you are concerned with, and in my above comments on frequencies for the cable lengths I used in the experiments the values are in the ballpark of those mentioned by GE.

As GE pointed out is is very difficult to match the line load end impedance to the characteristic impedance of the transmission line. Thus, other means are used to clamp or reduce peak voltage transients.

The frequency of resonance on the transmission line is controlled by line length and velocity of propagation. Not by line impedance. Line impedance in combination with load impedance will define the magnitude and phase of the reflected signal. Shorted vs open transmissions lines have a 2 to 1 frequency relationship relative to resonance.

.