Single-phase 120v to 3-phase 240v VFD to Leeson IEC motor

I've been searching the forums all day hoping not to bother everyone with a question, but I've hit all the dead ends I can see and don't know where to go next

I'm using this VFD:
KNC-VFD-CV20-1S-0002G

And this motor:
Leeson 192017.00

I'm unsure about many things, so I'll just load you with what I think I know and what I know I don't know.

The Problem:
I can power on the VFD, but when I set a frequency the motor makes a high pitch noise and the breaker trips.

Wiring the VFD:
The VFD is designed for a single-phase 120v input. From the unclear users manual, it seems L/N are L1/L2. I know some 120v VFDs use L3 as N, but this unit will only power on with hot at L1 and neutral at L2 (I checked the input voltage with a multimeter to be sure that wasn't the issue)

Wiring the motor:
I followed and triple checked the wiring diagram on the motor box for low-voltage. (Attached)

VFD settings:
I thought the issue maybe in the factory settings for the VFD being mostly 50hz. I changed all the 50hz defaults to 60hz and increased the max output voltage to the proper range for the motor low-voltage inputs.

I'm attaching images. Hopefully something jumps out.

Thise seem to come from Anaheim Automation. They do not look like the Huanyang units, so we'll assume they are somewhat legitimate.

Inverters that make high pitched noises like that are not happy. That noise is usually from a current limit. Possibly there is a short on one of the output wires, either to ground, or to another wire. To ground can be bad, since cheap low power VFDs are not always proof against that.

On the other hand, that does not usually pop the breaker.

The unit probably should show a display of Hz that varies as the speed control is turned. I assume it does not, and it may show a code instead. The codes are listed in the manual, showing what each means.

Disconnect the load wires from the VFD, then power it up. If it still does not give the right display of Hz, it may be dead.

Is there a chance that you hooked it up wrong at first, turned it on, but then corrected the error? If the input power is mistakenly connected to the output, that can destroy the unit. One symptom is to "throw codes" even though there is nothing connected to the load terminals.

Thank you for the reply. I really do appreciate the help.

Yes, I bought the unit from Anaheim automation. For now this is a small tester project, but it may turn into something bigger. Could you suggest a 1/4hp 120v single-phase to 240 3-phase VFD that is well made from a reputable manufacturer?

I do think that the issue may be in the motor wiring or an unseen short, which is obviously impossible for you to check through a computer screen .

The VFD does display Hz. When the unit powers on it flashes the Hz set point, which changes with the pot turn, until I press the 'Run/Stop' button. When the button is pressed, the power goes the the motor, the high pitch whine and slow turn begin, then the breaker trips. No error code on the VFD.

Let me know if any other info could be helpful.

As said... disconnect motor AT the VFD, then "run" it.

But, with no error code, AND the motor turning slow, it sounds like it is not doing the accel to speed. Not uncommon to hear a noise as the motor starts up. Make sure you have set everything you need to set.

Also, make sure all wires are making good connections. If one is open due to a bad connection or the like, it may do just as you say.

Usually, you need to set things like motor rated speed, motor rated frequency, motor rated voltage, max speed. The manual usually has a "quick start" section showing how to set up and get the motor turning.

JST said:

Thise seem to come from Anaheim Automation. They do not look like the Huanyang units, so we'll assume they are somewhat legitimate.

Click to expand...

Anaheim Automation is certainly legit. They don't make the VFD, don't claim to do, nor even have it "house branded". Buy-in, distribute & support, rather.

Kinco is no mystery, either, and no, Huanyang they are not. Had a German firm as partner, early-years.

Kinco - Milestones.

"Reputable enough" to carry-on here, anyway.

First thing I would WANT to do is try that VFD on another motor and/or the motor on another VFD.

Might not need to buy anything, just call a few bud's, toss the both of them in the car, go to where such goods are available and you can test.

Either unit is FUBAR, use your warranty or piggybank, not your lifetime-max time budget.

This is not a GFI breaker that is tripping out, not from overcurrent, but from the GFI, by any chance?

jim rozen said:

This is not a GFI breaker that is tripping out, not from overcurrent, but from the GFI, by any chance?

Click to expand...

Interesting thought. My Invertec manual says of ground faults:

Ground Fault Monitoring
As with all inverters, a leakage current to earth can exist. The Optidrive is designed to produce the minimum possible leakage current whilst
complying with worldwide standards. The level of current is affected by motor cable length and type, the effective switching frequency, the
earth connections used and the type of RFI filter installed. If an ELCB (Earth Leakage Circuit Breaker) is to be used, the following conditions
apply: -
 A Type B Device must be used
 The device must be suitable for protecting equipment with a DC component in the leakage current
 Individual ELCBs should be used for each Optidrive

Type B Device? Well, since I can't yet post an actual link, I can only suggest a Google search with the following words:

type b earth leakage circuit breaker

Gotta love companies who publishes stuff like this! Third one down, ABB...

On thing important to understand about "GFIs". GFI is actually not an official terminology any more, it's now basically a colloquialism that is too broadly applied and misinterpreted.

Officially in THIS country there are two types of ground fault protection: Ground Fault Circuit Interruptor (GFCI) and Ground Fault Protector for Equipment (GFPE). The levels of protection are very very different. Within GFCIs, there are different "Classes", A through D, defined by the level of protection as well and listed by UL. Despite the listing of these different classes of GFCI, when the NEC requires GFCIs for wet areas like bathrooms, kitchens, ponds, fountains and most recently, garages, that is ALWAYS referring to a "Class A GFCI", which means it will trip between 4 and 6 mA of ground current. For years, this was the only available class for residential use. Now recently, UL has been aligned with the IEC (European) standards that added classes B thru D. Anything other than Class A is, as far as the NEC is concerned, GFPE, it does NOT meet the requirements for GFCI set forth in the NEC for wet areas.

So when you read the manuals of a VFD made for worldwide use and they interchange the term "Ground Fault" with RCD (Residual Current Detector) or ELCB (Earth Leakage Circuit Breaker) of other terms they use, they are NOT referring to the same level of GFCI required here in North America. Class B for example is I think, 30mA of ground current; 5x the maximum allowed here.

There are no VFDs that will operate reliably behind a Class A GFCI without special attention to the installation or at the very least, having been tested together with specific parts. For example there are a lot of VFD powered pool pumps now, which MUST be protected by a Class A GFCI, so the pump mfrs have tested their units using a specific brand and model of GFCI breaker with specific wiring instructions. Any other brand or failure to follow all instructions results in nuisance tripping.

So, where does my breaker (which nuisance trips every time I put the Hitachi VFD to 'run') fit in this classification scheme?

Nomenclature from the breaker in question:

manufacture: General Electric
type: THQL-GFCI
"circuit breaker and ground fault cur interrepter, one pole unit"
120 volt 60 cycle class A CTL
UL listed, issue AC-980 E51075

And the trips were not so much nusiance, but rather, complete and absolute incompatibilty.

It is a class-A, meaning the MOST sensitive.

The only way you can hope to solve the problem is with heavy EMI filtering, most likely. And technique matters.

The GFCI trips if a current larger than the allowed current for its class goes out on a wire, and does not come back on the other conductor. The non-returning current is (justifiably) considered to have leaked to ground.

The way it is usually detected is a small toroidal transformer which both power wires go through. If the currents are balanced, everything cancels and there is no signal produced. If there is an imbalance, then the toroid responds to the difference current, producing a voltage output proportional to current. If that is large enough, the thing trips.

So far so good, but the electrical noise from a VFD can also create a signal that the sensor toroid picks up, possibly causing a trip. A good EMI filter will eliminate the noise signal, and avoid setting off the detector.

Unfortunately, many of them will also have capacitors that pass enough current to ground that the sensor may sense a legitimate leakage current, and can trip on that. So you need to be careful to select a filter that will not cause a leakage current issue. Many filters have ratings for that leakage, allowing you do select appropriately.

It is also possble that the VFD itself, or the VFD plus cable plus motor, can leak enough to be an issue. Then you are a bit stuck, but there are solutions to that also.

Since I have three phase at the shop, I only get into these matters when working on a friend's installation, but it seems to me that a motor's winding capacity to the frame would pass a considerable current at the high switching frequencies of VFDs.

Bill

9100 said:

Since I have three phase at the shop, I only get into these matters when working on a friend's installation, but it seems to me that a motor's winding capacity to the frame would pass a considerable current at the high switching frequencies of VFDs.

Bill

Click to expand...

It does, just common mode though. But with good grounding technique, the currents come right back to the VFD.

JST said:

It does, just common mode though. But with good grounding technique, the currents come right back to the VFD.

Click to expand...

That would mean floating (electrically isolating) the motor frame from the lathe. Tough to do.

jim rozen said:

That would mean floating (electrically isolating) the motor frame from the lathe. Tough to do.

Click to expand...

NO it does NOT mean that!

It seems counterintuitive, but it really is not when you look at the physics of the situation..

If you run the ground wire WITH the other VFD wires to the motor, bundled together, and especially if you use VFD cable, the high frequency currents will preferentially come back on the ground wire or shield, and NOT through the machine.

The reason is inductance, and the path of lowest energy. In order to go by some other higher inductance path, the currents have to "charge up" the inductance of that path, raising the impedance of that path vs lower inductance paths.. The path with the least physical space between the motor power wires and the ground wire will be the lowest inductance. A shielded wire will be the best at that, the shield around the wire is very low inductance compared to other paths, both due to the physical spacing, and due to the larger diameter of the shield vs wires..

If you do not PROVIDE a ground wire, then the currents have to return through/over the equipment, and you will have more radio interference, etc. The more widely spaced paths act as an antenna.

This is for the COMMON MODE current, which is generally very high frequency, since it is often due to the slight differences in pulse timing between the different VFD output wires. Common mode currents go out over the VFD wires, but return via a different path. Very similar to the currents that a GFCI detects.

All the "differential" currents are carried over the VFD output power wires, and do not contribute to the common mode currents nor the GFCI problem.

The inductance of the frame ground path is *much* lower than the inductance of the return path, in my case the
braided shield on the power lead to the motor, in parallel with the green wire inside the shield.

If the currents split by the inductance, most of that leakage will not go through the lead betwen VFD and motor,
it will go through the frame and green wire to the power source, via the lathe frame.

Sounds like I should break the sheild at one end, and put an RF choke in the green wire!

jim rozen said:

The inductance of the frame ground path is *much* lower than the inductance of the return path, in my case the
braided shield on the power lead to the motor, in parallel with the green wire inside the shield.

....

Click to expand...

How did you determine that the machine is lower inductance than the wires? Guessing? Rule of thumb? Clairvoyance? I would be betting you are dead wrong on that, unless you have some VERY good reasons.

The prime determining factor in the inductance is LOOP AREA. A path with more loop area is virtually certain to have more inductance in any but a very "pathological case". Especially since we are talking about currents well into the MHz region.

A return wire running closely parallel with the source wires will have a lot lower loop area than the machine, unless the wires are closely (very closely) following the surface of the machine, AND all the connections of machine parts along it are good.

The inductance of each of the two paths is not frequency-dependent. And the two paths
are in parallel, the current will split as the inverse of the inductance. Just physics.

This is easy to measure, I will do so and report back.

jim rozen said:

The inductance of each of the two paths is not frequency-dependent. And the two paths
are in parallel, the current will split as the inverse of the inductance. Just physics.

This is easy to measure, I will do so and report back.

Click to expand...

At some of the frequencies involved, the inductance of relatively small loops will be significant. The INDUCTANCE is not frequency dependent, but the IMPEDANCE is.

In any case, it works.... Running the ground reduces the EMI, which generally indicates that the current is redirected into the direct path. Of course it splits, the idea is to have the difference in impedance such that the split is advantageous.

No, it is NOT necessarily easy to measure accurately. It may actually be easier to calculate based on geometry than to get a good measurement.

There's no way to determine if the current division is advantageous or not, until you measure it.

Your hand-waving is about as good as my hand-waving.

Difficult to measure? How would you do it? I'll do it that way so you'll be less likely
to disregard the results if they disagree with your theory.

How would I do it?

1) use an inexpensive iductance meter, that measures at 400 hz. Float the motor on an insulator
and measure L between the motor frame and the green wire terminal on the VFD. Jot that number
down, and then mount the motor back to the frame and repeat the measurement.

2) same thing, with an HP LCR meter, measurement frequency variable up to higher frequencies.

3) borrow the high end network analzyer from work. A few GHz top end.