I need help choosing the right VFD for a 100 year old motor

Okay, I don't know if the motor is actually 100 years old but the lathe it is mounted on is!





The motor is 220V, 3HP, 3 Phase
At first I thought it would be easy to just "buy a VFD" but now that I'm actually looking I'm realizing there are a LOT more things to consider than I had originally thought... and now I'm getting overwhelmed with acronyms and options!

(I should mention that I will probably not ever use this lathe to it's full workload potential. This is my first lathe ever and I know it's probably overkill for a newbie like me.)

I have 220V single phase available right at the motor (this picture is before moving it to my place). I realize it's a big machine but to me 3hp seems like a lot for what I expect to ever do, is there a real danger in using a VFD that is "under powered"? The only reason I ask is that they seems to be cheaper as the HP demand goes down. If I'm not "pushing" the lathe will it not draw 3HP?

Does anyone have any advice or suggestions? Should I nix the VFD idea all together and just build myself a Rotary Frequency Controller? What about a static phase converter? While the lathe has gears to change speeds I like the idea of being able to infinitely control speed with a potentiometer.

I have lots more stupid questions where those came from!
Please help!
-Logan

A motor that old is not likely to survive being powered by a VFD for very long, although... a motor that old is questionable to begin with. You are well beyond the expected life of winding insulation, especially as was used in the early days of electrical manufacturing. You may want to look into having it rewound at a good motor shop and when you do, you can ask them to use Inverter Rated magnet wire.

If you do want to take a chance on it, you would be wise to add what is called a "sine wave filter" to the output of the VFD. It's not going to be inexpensive, but it is what I recommend for using old old motors that were never designed to be run from VFDs.

Jraef said:

A motor that old is not likely to survive being powered by a VFD for very long, although... a motor that old is questionable to begin with. You are well beyond the expected life of winding insulation, especially as was used in the early days of electrical manufacturing. You may want to look into having it rewound at a good motor shop and when you do, you can ask them to use Inverter Rated magnet wire.

If you do want to take a chance on it, you would be wise to add what is called a "sine wave filter" to the output of the VFD. It's not going to be inexpensive, but it is what I recommend for using old old motors that were never designed to be run from VFDs.

Click to expand...

I agree the motor may be questionable simply because of it's age, and at the moment I have no idea if this motor even works at all. It spins freely, but that is obviously not a testament to it's functionality. I have found this motor in a 1923 Westinghouse catalog so it could be at least as old as 94 years (it could be newer than that too, there's really no knowing).

If I choose to use a VFD (without a filter before the motor) and just use it until the motor gives up and dies am I asking for more trouble than preemptively replacing the motor or having it re-wound ahead of time? This is a home tool, I do not rely on it for income; as such it also will not be seeing production style run time.

Someone locally is selling 3HP single phase (220V) motor for $50... what would be the disadvantage to using a single phase motor vs a 3 phase motor if the HP is the same?
-Logan

I'm probably asking for flak now...

What about running the 3Ph motor on single phase power by wiring it in delta with a capacitor on one leg? I seem to recall seeing that done by an old timer somewhere...

It doesn't seem quite like the "right" solution, but it doesn't seem like a completely terrible option at this point either. It's certainly the cheapest of my ideas so far!

There is a little more to it than that. Starting capacitors need to drop out of circuit when running. Here is my solution on this old 15 HP

http://www.practicalmachinist.com/v...-and-vfd/how-make-old-15hp-3-phase-go-158528/

pillageTHENburn said:

What about running the 3Ph motor on single phase power by wiring it in delta with a capacitor on one leg? I seem to recall seeing that done by an old timer somewhere...

It doesn't seem quite like the "right" solution, but it doesn't seem like a completely terrible option at this point either. It's certainly the cheapest of my ideas so far!

Click to expand...

If you do not want to do the sine filter, at least an inductor in the output will help. Somewhere in the 2% to 3% area should be plenty good enough. It cuts down the spikes.

I guess inverter rated wire is nice enough.... but it really may not be that much better. Good varnish on the whole coil is probably as good in most cases. Old motors usually have better coil-to-coil insulation than the "slammed-together" new motors. Coil to coil insulation is more likely to have a fault than wire to wire, since voltages are higher, and any poor placement of insulation is a riisk of failure. The better wire does reduce the chance of two insulation defects being close to each other, I suppose..

An inductor on the order of 5% cuts out the voltage doubling effect, you need them when the transmission line between the vfd is of enough inductance to resonate with the capacitance of the motor.


Anyhow, if you set your vfd to 8Khz, any 5% load reactor can be used with delta connected capacitors at the motor, the vfd won't even notice the added line current due to the capacitors if they are not oversized. The motor windings will not have any dangerous high frequency voltages phase to phase. However, there is still the common mode voltage from all three phases to ground, for a 240v vfd its +170v, -170v at the carrier frequency of the vfd.

It is the common mode capacitance between all the conductors and the rotor, conducts current into and out of the rotor at say, 8khz. It is this common mode capacitance that shocks you when you touch an un grounded motor.. and when you ground the motor some of that current is going through the bearings.

To get rid of this common mode voltage requires an inductor with 5 legs, not three which is what most line/load reactors are. or use three separate ones.

Anyhow, the problem is you just can't use an inductor to get rid of that common mode voltage. In order to short out the common mode voltage you need Y connected capacitors, with the center of the y connected to the negative or positive dc bus of the vfd, and again you cannot do this with 3 leg "line/load reactors"


and yes you can run a motor from a vfd of lesser hp. reduce the volts per hz accordingly. at full nameplate volts the line current is on the order of a third of the full load amps at no load, and this would consume everything a 1 hp vfd could do, driving a 3 hp motor. at a 20% reduction in nameplate volts, the amps will drop by as much as half. for every motor, there is an optimal volts per hz that varies with the load.

at no load my 5 hp induction motor will run from 16 volts single phase, drawing 3 amps, about 40 watts power consumption.. that's the lowest volts it will run at. optimal volts for no load was something like 25 volts or something. no load at full nameplate volts per hz, power loss was something like 300 watts.

johnoder said:

There is a little more to it than that. Starting capacitors need to drop out of circuit when running. Here is my solution on this old 15 HP

http://www.practicalmachinist.com/v...-and-vfd/how-make-old-15hp-3-phase-go-158528/

Click to expand...

Yes, I was over-simplifying a bit. Your setup looks nice! Do you drop out all of your capacitors and run on only two coils of the motor? I feel like the one I saw had a "run" capacitor that was always in the circuit then a "start" capacitor which was wired parallel to the run cap but on a momentary (start) switch. Basically it "doubled up" the capacitors to get the motor up to speed then when you let go of the start switch it dropped the start cap and left the run cap...

Do you know how your motor is wired? (Delta or Y? My understanding is that this only really works with a Delta wiring?)
-Logan

There is only the start caps, all of which are dropped, Not using third winding of motor. Starts snappily, worked without issue since 2008

Too hot to go see about Wye/Delta

pillageTHENburn said:

Yes, I was over-simplifying a bit. Your setup looks nice! Do you drop out all of your capacitors and run on only two coils of the motor? I feel like the one I saw had a "run" capacitor that was always in the circuit then a "start" capacitor which was wired parallel to the run cap but on a momentary (start) switch. Basically it "doubled up" the capacitors to get the motor up to speed then when you let go of the start switch it dropped the start cap and left the run cap...

Do you know how your motor is wired? (Delta or Y? My understanding is that this only really works with a Delta wiring?)
-Logan

Click to expand...

To VFD or not to VFD...that is the question!

johansen said:

An inductor on the order of 5% cuts out the voltage doubling effect, you need them when the transmission line between the vfd is of enough inductance to resonate with the capacitance of the motor.

Click to expand...

If I go the VFD route I plan to mount it to the machine basically at the motor, so the run between the VFD and the motor would be no more than 2 feet max. Does this negate the need for an inductor? Most of this is at the brink of my knowledge but I thought I had read somewhere that a 5% inductor was needed in runs of 100' or more...

Anyhow, if you set your vfd to 8Khz, any 5% load reactor can be used with delta connected capacitors at the motor, the vfd won't even notice the added line current due to the capacitors if they are not oversized. The motor windings will not have any dangerous high frequency voltages phase to phase. However, there is still the common mode voltage from all three phases to ground, for a 240v vfd its +170v, -170v at the carrier frequency of the vfd.

Click to expand...

If I use a VFD is there still a need for adding capacitors to the system? I thought you could wire the VFD's 3-phase output directly to the motor?
If I don't use a VFD I thought I would try to run the 3-phase motor on single phase power using an appropriately sized capacitor to start it (like JohnO mentioned). I realize there must be some loss in this setup however as long as it's not damaging things I am okay with some loss.

and yes you can run a motor from a vfd of lesser hp. reduce the volts per hz accordingly.

Click to expand...

This can be done while programming the VFD? I think I generally understand the correlation between volts and current but hz is a new ingredient for me!

at full nameplate volts the line current is on the order of a third of the full load amps at no load, and this would consume everything a 1 hp vfd could do, driving a 3 hp motor. at a 20% reduction in nameplate volts, the amps will drop by as much as half. for every motor, there is an optimal volts per hz that varies with the load.

Click to expand...

So am I understanding you correctly in saying that if I use a 1hp VFD (with 220V single phase input) to power my 220V 3-phase 3hp motor I should limit the output of the VFD to 20% less than 220V (176V) - So I can set the VFD to have a maximum output limit of 176VAC? Do I need to worry about frequencies at this point?

Would this old motor be "happier" running at variable speeds being powered through a VFD or running using John O's capacitor start-up idea?

Thank you for being so patient with me while I attempt to wrap my brain around this!

To some extent if the motor is located nearby the vfd then the transmission line inductance is too small, pushing the resonant frequency higher than the harmonics the vfd can practically generate, and so there is no voltage doubling spikes going on. In practice there always is a voltage spike generated, and it could also occur inside the motor itself.

If I use a VFD is there still a need for adding capacitors to the system? I thought you could wire the VFD's 3-phase output directly to the motor?

Click to expand...

you're confusing why i'm talking about delta connected capacitors after a line reactor, vs using capacitors to start a motor from single phase power.

Regarding a traditional 3 leg "line/load reactor" and delta connected capacitors.. it doesn't take much inductance to smooth out the phase to phase voltage. however, you still have the common mode voltage (all 3 phases rising and falling together, relative to the dc buss of the vfd, at the pwm frequency) that is what destroys bearings.


You may have to set the vfd as low as 57% of nameplate volts to get a 1 hp vfd to drive a 3 hp motor and to actually get 1 hp from the motor shaft without over loading the VFD. (Usually vfds can handle more amps than the nameplate amps that are "nominal" for the same size motor, so you can increase the volts as high as you can go before you overload the vfd. the number of 57% comes from the square root of the ratio of the hp. but perhaps you could go as high as 70% according to the current capacity of your "1" hp vfd)

So you would program the 1 hp vfd for 140volts/60hz, even though the motor is a nominal 220v/60hz motor.
you will need to lower the percentage of boost volts to avoid over current at turn on. This is because a 3hp motor is far more efficient than a 1hp motor. so if its default is 20%, try 10%.


anyhow, you can run that motor as if its a split capacitor motor without a vfd. It will be safe to pull about 1 hp from it, but a vfd gets you infinitely variable speed.


but i have no idea how long sleeve bearings last when the motor is run from a vfd.. (does it have sleeve bearings?)


anyhow one of the options available is this.. AEGIS SGR-.875-UKIT Bearing Protection Ring, Dia. 7/8 In | eBay

i had no idea they were that expensive.

its a whole bunch of microfiber brushes that short the motor shaft to the frame, so that the common mode current generated by the pwm harmonics acting through the capacitance from the motor windings to the rotor, go through the microfibers instead of through the bearings.

it might be worth trying to just install a regular carbon motor brush and short the shaft to the motor frame that way, but you'd have to replace it every 3000 hours or twice a year...

johansen said:

but i have no idea how long sleeve bearings last when the motor is run from a vfd.. (does it have sleeve bearings?)

Click to expand...

Based solely on its age I believe it has sleeve bearings but I haven't looked closely yet. I found an interesting article about shaft currents and it states:

If the shaft to frame voltage exceeds 100 millivolts AC for ball or roller bearing, or 200 millivolts AC for a sleeve bearing, the shaft current is probably high enough to degrade the bearings.

Click to expand...

The full article can be found here: EASA-Shaft-Bearing-Currents.pdf.

johansen said:

anyhow, you can run that motor as if its a split capacitor motor without a vfd. It will be safe to pull about 1 hp from it, but a vfd gets you infinitely variable speed.

Click to expand...

I really like the idea of variable speed but I might opt for the cheaper option of powering it from single phase with capacitors for now. The way John Oder wired his (10HP) motor the capacitors are only in the circuit for starting, what are your thoughts on keeping at least some capacitors connected while running? It seems to me that, while not extremely efficient, it might be more efficient to keep a capacitor linked while running instead of dropping out completely. Is that incorrect?

I did not create this diagram but this is basically what I had in mind. The start switch is momentary so it is held down until the motor reaches speed.


Thoughts?