New VFD's on motors

Status
Not open for further replies.
GoldDigger said:
When you have a transmission line (looking at the characteristic impedance rather than the bulk inductance and capacitance separately) you can look at reflections that occur at impedance changes........
Click to expand...

Never used a 60Hz TDR, but....

A while back, I spent time pondering this topic. And have a question to add to the muddle.

What happens if the load reactor is mounted at the motor,
X hundred feet from the VFD?​

Discuss.
 
Open Neutral said:
Never used a 60Hz TDR, but....

A while back, I spent time pondering this topic. And have a question to add to the muddle.

What happens if the load reactor is mounted at the motor,
X hundred feet from the VFD?​

Discuss.
Click to expand...

My guess is it still provides some protection to the motor, but the reactor will take the hits from the high voltage spikes. Long conductor length and wave reflection or something like that is the problem and putting the reactor near the drive trims those spikes before they get compounded by the conductor length.
 
Open Neutral said:
Never used a 60Hz TDR, but....

A while back, I spent time pondering this topic. And have a question to add to the muddle.

What happens if the load reactor is mounted at the motor,
X hundred feet from the VFD?​

Discuss.
Click to expand...
The "spike" issue is the result of the fast rise time of the leading edge (dV/dt) of the PWM DC pulses causing the conductors to act like capacitors, and a charge builds up between them. That charge gets reflected back and forth between the drive and the motor until it builds to the point where its potential exceeds the insulation rating on the magnet wire in the windings or the cable insulation, sometimes both. Putting a reactor at the drive adds an inductive time constant to the PWM pulse current, because in an inductive circuit, the current through it cannot change instantaneously. That then also slows the voltage rise time as well, reducing (not eliminating) the cause of the reflected wave phenomenon. Putting the reactor at the load end can allow the reactor to absorb those voltage spikes before they reach the motor insulation, but do nothing to protect the cable.

Using the shielded cable (as Besoeker does and as do most people in the EU because it is required) significantly reduces that cable capacitance issue, that's why he rarely sees problems. In this country, THHN gets pulled for motor leads because that's the way it was always done (hopefully in steel conduit), but that does nothing to reduce that capacitance issue.

The most common reasons why people often say "I've never had a problem" are:

1) most of the drives they have done are 208-230V. The issue exists, it's just that the spikes don't exceed the insulation rating.

2) they have been using older drives that still used darlington transistors, which have a much slower turn-on time, so the leading edge of the PWM pulses is not as steep.

3) they have been using newer IIGBT based drives, but ones that employ specific circuit design techniques to reduce the capacitance and thus increase the acceptable drive-to-motor distance. Case in point, many AB PowerFlex drives list the TESTED drive-to-motor distance as being hundreds of feet, with no load reactors, filters, or shielded cable. Adding any of those extends that distance even further, but at that point, voltage drop becomes problematic too so people tend to keep the distance short anyway.

4) they are always dealing with short distances, as in under 50ft.
 
GoldDigger said:
When you have a transmission line (looking at the characteristic impedance rather than the bulk inductance and capacitance separately) you can look at reflections that occur at impedance changes.
With the large wire size of non coaxial conductors the characteristic impedance will be low compared to the motor load reactance. That means that any impulse (sharp step in voltage) will be reflected back in the same polarity as the incoming impulse (since no current will flow at what we approximate to be an open circuit termination.) That means that as a first approximation the step voltage at the motor will be twice that of the input at the other end of the cable.
This voltage will persist until the reflection has had time to be absorbed or re-reflected at the VFD end of the cable. Hence the longer the cable the longer the voltage spike will persist.
The source end at the VFD looks like a short circuit compared to the characteristic impedance, so the reflection from that end will have opposite polarity and will try to cancel out the double voltage when it gets back to the motor end.
You can also look at the process in terms of harmonics (not of 60Hz but of the switching frequency) which can be very high harmonics indeed. And then look at standing waves at those frequencies. But to the extent that the wires can be considered as a transmission line, the attenuation over distance of those high harmonics will be small.
An RF coax cable can pass 1000 MHz without too much trouble....
Tapatalk!
Click to expand...

Pictures worth a thousand words? We put this approx. 150ft long vfd shielded cable into use about 20 years ago. It burnt from the standing waves in about 3 months time. Looks to me like the approx. 5" apart burns give a standing wave at around 300/.3meter=100mhz, assuming both positive and negative spikes burnt alike so a burn mark every 1/2 wavelength. IIRC we 'fixed' this one by changing out the cable to another insulation material that would not burn (and no longer clear)! Fixed the symptom not the problem, but that system is still running today... Not sure we knew about standing waves (other than in our ham radios) back then, just the transistor switching spikes....

https://www.box.com/shared/hrrsr88q1sysw5tvxlvi
 
Jraef said:
Using the shielded cable (as Besoeker does and as do most people in the EU because it is required) significantly reduces that cable capacitance issue, that's why he rarely sees problems.
Click to expand...
I have seen problems with motor winding failures (the PD effect) but not with bearings. The absence of bearing failure experiences is, I think, because of the type of cable.
 
Jraef said:
Using the shielded cable (as Besoeker does and as do most people in the EU because it is required) significantly reduces that cable capacitance issue, that's why he rarely sees problems.
Click to expand...


This is where I get confused. I think of twisted pair as a linear low pass filter. Long telephone loops lose the higher frequencies; that's why DSL is not possible on longer ones.

Shielding is a separate issue but increases the cable's capacitance, but to ground.

My SWAG thus far is this not "shielded cable" as much as "low capacitance cable for VFD's that is also shielded..."

And I infer the cable fires result from cable that is too capacitive aka leaky & draws current.

What I have trouble with is: to me the longer the run, the more rolloff of HF components, and the less grief. I'm thinking now that the runs are too short [in wavelength terms, not feet] to mitigate the standing wave issue, but long enough so there IS a standing wave problem.

Who'd a thought I need a 60-180Hz slug for my Bird wattmeter?

<snark>
Can't we just install waveguide between the VFD & motor, and solve the issue?
 
Status
Not open for further replies.
Top