VFD question

Indeed they do.

I have one (overpowered -- 7.5HP) VFD which I have used this way (probably about a 30 foot cable to a 1HP motor -- shorter for the 2HP Bridgeport motor), but I am working towards having a dedicated VFD for each machine. The reasons for this are:

1) I can switch the VFD from the original motor drum switch on the machine, so the feel of operation is not changed. 2) On the Bridgeport, (which Is CNC, being modified to run from a linux/EMC controller with servos instead of the original LSI-11 one with steppers), I will be able to feed an axis signal to the VFD to have the computer control the spindle speed -- other than requiring the engagement of the back gears, or changing the vari-speed pulley for extremes of speed or torque requirement.

Note that because the VFD is so over-capacity, I can get away with switching the motor leads between the VFD and the 1HP motor on my Nichols mill, something which would risk damaging a more reasonably-sized VFD.

Ideally, if you are going to be running a lot of machines, and don't really need the variable speed feature, I would suggest that you build a rotary converter instead, and route that to all machines. This is cheaper than using an individual VFD for all of the machines, and still allows you to put an indidividual VFD between the rotary converter and any single machine which needs variable speed. Since you apparently don't have these yet, consider that at least some Bridgeports come with a variable-speed pulley assembly, and serious bandsaws (with three-phase motors) are likely to also have mechanical variable speed built in, so you don't need that from your VFD. The South Bend lathe, however, could benefit from the variable speed function, so *that* I would put on a dedicated VFD.

I'm not sure whether having variable speed for the buffer will be a benefit or not. Something to be determined -- in part by what materials you are buffing. Plastics would probably benefit from the variable speed, but metals maybe not

Enjoy, DoN.

In reality it's a giant pain in the ass.

You've got extension cords draped all over and if you are going back and forth from the lathe to the mill, you drive yourself crazy.

It might just be me, but when I'm building something I get real focused on the task at hand and it drives me nuts to have to do some silly thing like change plugs. Your speed and direction control will probably be at the vfd, 20 feet away, nullifying the basic beauty of having the vfd.

I bought 2 cheap ($100) 1 hp VFD's and mounted one each on my lathe and mill. Works great and you'll probably end up spending $100 on power cables and decent twist-lock connectors.

Baldor disagrees with you.

Since I just got a brand new 3 phase motor for my milling machine, and I = already have a 3 phase motor on my lathe (they are both belt driven and constanly=

changing the belts drives me crazy), I have looked into the same issue al= so. =

This is what Baldor says about this:

Lead Lengths The best installation will keep inverters and motors close together. The effect of reflected wave voltage is generally not a problem if the power run is less than 15 feet (5 meters). However, as the run gets longer, voltage at the motor terminals rises higher than the insulation system=92= s design voltage. One installation had 30 motors driven from one inverter. Although the first motor saw 460 volts (RMS), the last motor,

1000 feet of wire away saw 2000 volts. The best way to solve this problem is to not create it in the first place. Keep the run between an inverter and motor as short as possible.

And:

Installation =02 Distance between the motor and inverter should be minimized. If they = must be very far apart, specify a load reactor and a low pass filter in between. =02 Monitor the temperature of the motor. A simple thermostat in the moto= r can prevent catastrophic failure from overload or too low a speed. =02 Grounding is very important. The motor and inverter should be at the = same ground potential and this common ground should be tied to true earth ground. Failure can occur= from the motor and inverter at different voltage references. =02 Load reactors can also be used as line reactors. Sometimes other equi= pment in the building (like compressors) or power company capacitor bank switching, can introduce lin= e voltage spikes which may cause an inverter to shut down on overvoltage. Line/load reacto= rs can help reduce the magnitude of the spike and let the drive continue operation.

The entire article titled: "Fundamentals of Inverter=96Fed Motors" can b= e found here

Good luck

You also need to hook up the start/stop switch to the VFD, since the motor leads should go from the VFD output directly to the motor. The wiring involved is easy, but it is another cable needed from the machine to the VFD. Running a machine with no stop switch within reach would be a foolish risk even as a temporary measure.

VFDs are essentially closed-loop feedback devices. Long runs introduce inductance that interferes with that feedback. The result is high- frequency overvoltage transients at the motor which can break down the insulation and short out the windings. Not good.

The latest AB vfd's (Powerflex 700) that I'm using include software correction to limit the reflected wave overvoltage to 2X nominal up to

300 ft. Inverter duty motors are built to stand over 3x. Line reactors or other protection can extend the allowable length. Luckily for me, none of my runs, at home or at work, are longer than about 10 ft., so I don't worry about it.

All that aside, I agree with the previous post which stated that switching cords would just be too big a PITA. Also, vfd's really don't respond well to switching between the vfd and load unless they're powered down, leading to the possibility of screwing up and forgetting to power down (powering down to switch would be another PITA).

By the way, most of the AB documentation and manuals are available on-line. Go to

and look for the A-Z product directory. From the product pages, there are links to the on-line documentation.

Pete Keillor

Yaskawa(Magnetek) and Mitsubishi manuals are also available online.

Ned Simmons

The baldor VFD I have at work drives a 1/2 hp baldor motor, through a line length of about five feet. The VFD does periodically fault out, with a 'general protection' fault - nothing specific.

The manual suggests that this behavior may be from ground faults inside the motor (which I cannot find, with an ohm meter) so I am wondering if maybe the 5 foot line is too capacitive and causing the fault conditions.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Just a thought, does the vfd have communications capabilities? I found the low Hz - high current shutdown timer I mentioned earlier by watching the vfd operate with a laptop and AB's serial hookup, the same thing I use to program them. That might help. That model AB has over 300 parameters, so there's a lot to look at.

Good luck.

Pete Keillor

One thing to keep in mind about standing waves, the line has to be at least 1/4 wavelength electrically for a standing wave to form. For

60 Hz, that's a distance of 1,250,000 meters. That's why power companies don't have to worry much about standing waves on their lines due to the loads you attach to the line.

Now since the VFD produces a stepped sinewave, you need to look at the switch frequency and harmonics too. Generally, odd harmonics through the 5th are considered significant. So lets consider a 4 kHz switch rate, giving a 5th harmonic of 20 kHz. A 1/4 wave at that frequency is 3750 meters.

Of course the line is terminated in a reactive load (the motor), and that reactance will electrically lengthen (inductive) or shorten (capacitive) the electrical length of the line. There are standard transmission line formulas for determining by how much a given load capacitance or inductance will shorten or lengthen the line. But suffice to say that the reactance would have to be *huge* to cause a reflected wave problem on a

20 foot line.

Gary

Some newer VFDs use a pwm scheme, and the spikes have what seem to be microsecond risetimes under some circumstances.

That's going to get fourier components up near the point where they could be a problem - I think that's what the Baldor tutorial is getting at.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

That's funny, so does my Windows machine...

On Sat, 04 Oct 2003 17:00:21 -0700, Jim Stewart pixelated:

His VFD must be running Windows for Wocket Scientists.

.-. Life is short. Eat dessert first! ---

Comprehensive Website Development

The cable to your keyboard is too long then....

:^)

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Sure thing, so do the stepped sinewave units. But a good design will have low pass filtering on the output to smooth those sharp edges down toward the sinewave shape that the motor really wants to see.

Gary

The Baldor ones don't do that in the VFD. They require you to puchase a 'load reactor' which is nothing more than a large three phase choke to install between the vfd and the load.

I think their thoughts are, most installations don't need them, it's an extra cost to install them, so they make them offboard as an add on item.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Where did you buy these?

Gary Repesh