Teach me about proper VFD engineering (line reactors)

The title pretty much says it all. I have very little experience with VFD's, but it's something I want to learn more about (primarily because I have a project coming up where I think one will be of good use, that is to vibrate a jet-ski in a themed environment to simulate that it's running (poorly)).

My only experience previously was to drive a "vortex tunnel" in a haunted attraction (here's a good video of a somewhat typical build and the effect that you get from it; https://www.youtube.com/watch?v=9_jnKBIvTK4)

Originally mine was driven with a gear motor, but due to a flood it needed to be replaced. So I replaced it with a 1/2hp motor and GS1 drive from Automation Direct. 1ph 120v input, 3ph output. No fuses, no line reactors. I had a few simple pushbuttons wired up for start and stop, as well as some E-stops in place. I learned a little about ramping speeds and such. Basically, I was able to experience the minimum of what one should know about VFD's.

Now I'm going to use one with a off-balance weight to produce some vibration for the jet-ski. I'll probably use a 1/4hp motor and a GS2 drive. The motor won't run often, say 5 minutes once an hour at best.

My question is, how important are line reactors? My previous setup for my haunted attraction ran for ~5 hours a night for 32 nights straight for 4 years now. That's not to say that I don't need them, but that setup has worked flawlessly. I guess at this point I'm trying to learn a bit more about the engineering side of it, as well as the control side of it. I'll be using a P2K to control it over their GS Drive bus.

Here is a good whitepaper from Allen Bradley:

Been stung enough times where a line rector was needed to fix flaky behavior but not designed in; I don't hesitate to use them in but I'm not always the circuit designer. Got a few customers that are on the ball and insist on them.

A line reactor is cheap insurance for the VFD. In a gross simplification of a complex concept, it's main role is to add impedance and inductance to the circuit feeding the drive. The drive's rectifier feeds energy into a DC bus that is supported by capacitors, then the inverter pulls energy off of that bus to send to the motor. The rectifier diodes don't conduct continuously because they have a Forward Conduction Voltage (FCV) threshold, so current drawn by the diodes is done so in gulps of power, also called "pulses" (hence the term "6 pulse inverter"), only at the peak of each input voltage sine wave.

If one of the incoming line sine waves has a severe voltage transient on it and the sine wave never gets to or drops below the FCV threshold, that diode does not conduct (contribute) into the DC bus. If the VFD is running a motor, the inverter transistors are sucking energy out of the DC bus, draining the caps. But if the caps are not replenished by the non-conducting diode(s) in the rectifier, even though for only a fraction of a second, they will recharge themselves at the next available opportunity, namely the NEXT diode pulse. Because caps charge instantly, they will draw that charging current at the AVAILABLE FAULT CURRENT level, however briefly. The amount of current they draw can (eventually does) harm the diodes and the capacitors themselves. This by the way also takes place when you initially power up a drive, but we use a "pre-charge" circuit to slow down that current rise. The pre-charge is out of the circuit while it is running however.

What the reactor does then is to add in an "inductive time constant" effect to the circuit feeding the rectifier diodes. The inductive time constant describes the fact that in an inductive circuit, voltage and current cannot change value instantly, it has a prescribed rate of change. So what the reactor accomplishes is slowing down the rate of rise of current in that attempt by the caps to replenish themselves through one diode, stretching the energy transfer out over a few cycles and avoiding stressing the components. In a lab setup I participated in, we connected a 25HP drive to a 100kVA transformer fed from the grid. Every time we detected a grid transient, likely from a grid switch by the utility, that little 25HP drive rectifier drew one or two pulses of 805A, as opposed to about 40A normally. Adding a 3% line reactor knocked those pulses down from 805A to 55A. That drive will survive a lot more abuse from the utility with that line reactor.

A transformer would do the same thing, but a reactor is less expensive and smaller. But if you DO have a drive fed by a transformer that is less than about 10x the kVA of the drive (some say 20x, I am conservative), you may not need the reactor.

Reactors also slow down the change in values of harmonic currents as well, hence their contribution to reducing harmonics to a small extent. But in general just a reactor alone doesn't really "solve" a harmonics problems, it just lessens it. An inductor on the DC bus has about the same effect, but does not offer the same protection against line disturbances as the reactor, because it is AFTER the rectifier.

jraef said:

A line reactor is cheap insurance for the VFD. In a gross simplification of a complex concept, it's main role is to add impedance and inductance to the circuit feeding the drive. The drive's rectifier feeds energy into a DC bus that is supported by capacitors, then the inverter pulls energy off of that bus to send to the motor. The rectifier diodes don't conduct continuously because they have a Forward Conduction Voltage (FCV) threshold, so current drawn by the diodes is done so in gulps of power, also called "pulses" (hence the term "6 pulse inverter"), only at the peak of each input voltage sine wave. <snip>

Click to expand...

AWESOME. You put it in words that I can fully understand. I suspect a reactor also helps with inrush current as well?

My next question, if I'm using a 0.5hp drive and a 0.25hp motor, should I size the reactor for the motor or the drive? In my head, I would say motor, but the tech at AD said to match it with the drive. But, he also seemed to be reading more from a script that "if using a 0.5hp use 0.5hp reactor".

Brandon_K said:

AWESOME. You put it in words that I can fully understand. I suspect a reactor also helps with inrush current as well?

My next question, if I'm using a 0.5hp drive and a 0.25hp motor, should I size the reactor for the motor or the drive? In my head, I would say motor, but the tech at AD said to match it with the drive. But, he also seemed to be reading more from a script that "if using a 0.5hp use 0.5hp reactor".

Click to expand...

There is no inrush current when using a VFD. The reactor would make no significant difference anyway, for example on an across the line starter. The reactors we use for VFDs have only a small effect on limiting current, they just extend the time it takes to get there which for the VFD, means the energy is stretched across more diode pulses. There are "current limiting reactors" used for that purpose, but they have much higher impedance and thereby create a more significant voltage drop so you don't want that with your VFD since you don't need it. A 5% reactor is about all you might ever need, 3% does the job in most cases.

You can size the reactor for the motor, but it's not a good idea because if, in your case, you someday want to repurpose that drive in another machine that has a 1/3 or 1/2HP motor, you will have to buy another reactor. The difference in price is negligible, in that size likely only a few dollars. I would recommend buying the reactor to fit the drive.