What's new
What's new

Problems using 220v VFD with step up transformers to power Hardinge HLV Lathe

rickbrennan

Plastic
Joined
Jan 19, 2018
Location
Boulder, Colorado, USA
I have a beautiful older Hardinge HLV (not HLV-H)that I got out of the back room of a university. Very little use on it. Runs on 440V. I've read tons of forum posts about different options for powering the lathe from 220v single phase power and settled on an Automation Direct GS2-23PO VFD (3HP rating) to manufacture the third leg and connecting the VFD output to three 1KV 440/220 step down transformers in reverse for step-up output. The transformers are wired in a delta-delta configuration. I tested the transformers individually before wiring them together and they all test good (220 in, 110 out). When I hook them up to the VFD with the disconnect at the lathe tripped (motor not connected), I immediately get an overload condition and resulting error and shutdown on the VFD. Tech support at Automation Direct wasn't familiar with this use case and couldn't help me diagnose the problem.

I don't want to experiment with this setup and risk smoking the relatively expensive VFD and hope someone out there is knowledgable about connecting transformers to VFDs and can help me think through what I've done wrong.

I've read about different wye and delta combinations. They describe the wiring and how to calculate things like phase current and phase shift, but I've not been able to read anything about running VFD waveforms through transformers and what considerations for the configuration options might be. For example, do I need to use a delta-wye setup instead of delta-delta? Or, is there a problem with inrush current running step down transformers in a step-up configuration? If so, are there settings on the VFD that would make this system work smoothly?

Also, if the delta-delta setup is ok, how do I test the transformers wired together to ensure I'm getting the desired behavior before I connect the VFD?
 
VFD's 101:
.
Do NOT run the VFD without a motor connected to the Output.
Should say that in more than one place in the manual.
Go have a look on page 2-11 of the manual that shows the connection diagram.
Do you see any disconnect BETWEEN the drive and motor?

Pretty lame tech support at AD didn't tell you this right off.
:rolleyes5:
 
Does the VFD have an auto-tune feature? It would seem to me that you would need to get the current loop tuned before trying anything. Also, set up the current limit on the drive to 50% of capacity when starting out and work your way up from there.

You shouldn't need to connect the motor if you have a transformer between the VFD and the motor. Check the resistance at the input of each transformer also - it should be roughly half the resistance of your lathe motor if of generally the same power rating.
 
Convention is to step up the voltage before the VFD, and then to run a 440V VFD with a direct connection to the motor. Typical transformers are made to operate at a given line frequency, and cannot be fed effectively with the VFD variable output and they are designed for true sinusoidal wave forms. They will heat up and fail otherwise. I do not see this as a fault of AD unless they told you that the VFD could be used in that manner, which is doubtful. You also can't just feed the lathe the power from the VFD and expect all the controls to work. Doesn't work that way, if you wanted to just generate the third leg and have that power the machine, you need a RPC or a Phase Perfect VFD.

There is some discussion that I have seen for Yaskawa VFDs that use an output reactor followed by a step up transformer, it gives some general guidance if you want to persist with this approach. What I have seen discussed is that the VFD needs to be 2X the rated motor output, and probably should be run in a straight V/Hz at a low carrier frequency. You still would need a VFD output reactor before the transformer.
https://www.yaskawa.com/delegate/ge...D.08&cmd=documents&documentName=AN.AFD.08.pdf
 
you should have no problem running a transformer from a vfd, but you must program the vfd for linear volts per hz, otherwise you will saturate the transformer at turn on. this is most likely the reason you were un successfull.

the issues with pwm and leakage inductance and the capacitance of the transformer windings are legitamate, personally i would monitor the secondary transformer voltages with an oscope to verify you have mostly nice sine waves and no spikes higher than 700-1000volts.


you will likely find the transformers to have enough leakage inductance you can install line to line capacitors after the transformers to get clean sine waves, the transformers will run hotter and you will need to verify they have the headroom to handle the heat. if they don't then install a line reactor as mksj said.

but again, linear volts per hz, starting at zero. no boost volts. no vector control.. or you saturate the transformers. alternatively you could use a higher voltage transformer, but when the vfd sends 30 volts AC at 5 hz to start the motor.. that's equivalent to 360vac/60hz. and it might try to send more than that, 20 volts at 2 hz, or 30 volts at 3 hz, you would need a 480v/960v transformer to avoid saturating at startup.
 
As mentioned, the PWM voltage waveform will not be optimum for the transformer. We have put a sine filter at the output of the VFD . . . and we have run motors with encoder feedback accomplishing positioning through multiple transformers in sea based tidal generators. VFD -> sine filter -> 480:4160 transformer -> 2.6 km undersea cable -> 4160:480 transformer -> PM motor (generator) with encoder . . . as such take the linear V/Hz recommendation with a grain of salt if you want to get sophisticated with commutating motors or generators at low speeds (<4 Hz).
 
As mentioned, the PWM voltage waveform will not be optimum for the transformer. We have put a sine filter at the output of the VFD . . . and we have run motors with encoder feedback accomplishing positioning through multiple transformers in sea based tidal generators. VFD -> sine filter -> 480:4160 transformer -> 2.6 km undersea cable -> 4160:480 transformer -> PM motor (generator) with encoder . . . as such take the linear V/Hz recommendation with a grain of salt if you want to get sophisticated with commutating motors or generators at low speeds (<4 Hz).

you could probably push the volts per hz up to 50% higher than the transformer was designed for starting purposes only, but out of the box the initial default 20% voltage at zero hz is going to blow the over current. yes, that's 48 volts at like 1hz. some vfds may default to 10%, you can go read the manual.

P2.07 is 10 volts.
P2.06 is 1.5hz.

reduce 2.07 down to 6 volts and try again.
 
No need to boost the V/Hz when starting a PM motor . . . my point is that as long as you keep the flux density the same, and soften the PWM with a sine filter, you can run at any frequency you want without issues with the transformer. In this case the drive was configured for Field oriented control rather than linear V/Hz.

No cowboy antics here, I have both participated in the design of PWM inverters as well as written software to control them and read my share of manuals and white papers as well as written a few.
 
A litle more background: I don't have 3 phase power available in my shop, only 220v single phase. I can step this up to 440v single phase with a transformer before the VFD, but don't see 440v VFDs designed to be powered with single-phase power. That's why I started out with a 220v/single phase input VFD. The motor is rated at 1.5hp and I purchased a VFD specified for 3hp to handle any inefficiencies in the transformer setup.

I intend to power the motor from the VFD and wire the controls separately to the control side of the VFD to take advantage of the VFD functionality.

The motor on the Hardinge lathes of this vintage is specialized and application specific. Not easily replaced. I'd like to get input on how to feed this 440v three phase motor from 220v single phase shop power in a way least likely to damage the motor.

I'm not wedded to the current setup, rather looking for the best way to solve the problem of powering a 440v three phase motor with 220v single phase power and utilizing all the available VFD options (variable speed, braking, jog, reverse, constant speed under varying load, etc).

Sorry if my original post wasn't clear.
 
You are using three transformers to form a three phase connection? Got a picture of that?

Get a 220v three phase motor and get familiar with your VFD.

Then maybe a 3 phase transformer with one iron core.

If you wire the motor direct to the VFD/transformer and it works there is the issue of the lathe cabinet controls. It was built for 440v.
 
Rick,
I've powered several 440V Hardinge machines, including an HLV-H from 230V single phase. Transform 230V single to 460 V single then power a VFD for motor functions. In each case, the VFDs did not state that they would operate on single phase input but careful reading of the manual or a call to mfgr's tech service confirmed that it would, including which input terminals to use and derating required. Both with Allen Bradley and ABB VFDs.
 
A litle more background: I don't have 3 phase power available in my shop, only 220v single phase. I can step this up to 440v single phase with a transformer before the VFD, but don't see 440v VFDs designed to be powered with single-phase power. That's why I started out with a 220v/single phase input VFD. ...
Technically, any VFD is capable of accepting single phase input and giving you 3 phase output, it's just that some mfrs don't let you do that. That's because on a lot of 480V rated drives, they don't expect that someone will connect single phase, so they build in a Phase Loss Detection circuit, then don't give you a way to defeat it. the drive COULD work fine, they just didn't have the foresight to allow you to do that. Drives like this have however been falling by the wayside because it limits their marketplace, so most new drives on the market now are fine with single phase input at 480V.

In the conversion process, all the VFD is doing with the input AC power is converting it to DC, so it doesn't really "care" how many phases come in. The DC that comes off of the rectifier is "pulsating" DC, which then goes to capacitors to smooth it out for use by the transistors to make the Variable Frequency / Variable Voltage output. But with a single phase input, there will be more "ripple" on that DC bus, so it needs more capacitance to smooth it out. As a gross general rule then if you de-rate a VFD enough, it will work fine.

HOW MUCH of a de-rate is dependent on the internal design of the VFD. Some of the industrial grade drives made by US based manufacturers will have what's called a "DC bus choke" and therefore will require only a 50% de-rate, meaning you buy a VFD rated for 2X the current as the motor nameplate. But smaller cheaper drives that only use capacitors typically need a 65% de-rate, or they need to have a temperature de-rated, meaning 25C (77F) operation instead of 40C (104F), which for most industrial applications is unobtainable without air conditioning. People have used 50% for years though and gotten away with it, but the newest generations of smaller-cheaper-faster drives can't take it like they used to.

So if you buy an inexpensive drive like AutmationDestruct or any other low cost Asian variety, I recommend a 65% de-rate (motor FLA / .35). If you find an American or European based design that says it has a DC bus choke, then use a 50% de-rate.
 
Technically, any VFD is capable of accepting single phase input and giving you 3 phase output, it's just that some mfrs don't let you do that. That's because on a lot of 480V rated drives, they don't expect that someone will connect single phase, so they build in a Phase Loss Detection circuit, then don't give you a way to defeat it. the drive COULD work fine, they just didn't have the foresight to allow you to do that. Drives like this have however been falling by the wayside because it limits their marketplace, so most new drives on the market now are fine with single phase input at 480V.

The only thing that can be tried is to jump one of the incoming lines to the third phase input. If the drive is designed to only detect a
potential on the line then the drive won't fault. If the drive is measuring zero crossing times on each phase then a jumper will not work.

Added opinion:
The drive has a lot more important work to do than to check that all lines are 120 degrees apart. But it isn't difficult just wasteful of cpu time.
Would use three microcontroller input pins and service three interrupts and verify that they occur at 120 degrees. Probably this kind of thing can
be done every 30 seconds and not be much of a bother to a fast processor.

Zero-Crossing Detectors Circuits and Applications
 
Last edited:
Given your transformers are connected delta-delta, I wonder if one of the windings is reversed. That would mean it's essentially trying to drive into a short circuit.

Try disconnecting one of the corners on the secondary, so you have an open delta.

I think you should be able to treat it the same as on a 6-lead motor; fire the drive up and check if you have significant voltage across the two open points. If you do, one of the windings is reversed. Try reversing each one at a time; if that winding doesn't fix it, put it back to how it was.
 
The only thing that can be tried is to jump one of the incoming lines to the third phase input. If the drive is designed to only detect a
potential on the line then the drive won't fault. If the drive is measuring zero crossing times on each phase then a jumper will not work.

Added opinion:
The drive has a lot more important work to do than to check that all lines are 120 degrees apart. But it isn't difficult just wasteful of cpu time.
Would use three microcontroller input pins and service three interrupts and verify that they occur at 120 degrees. Probably this kind of thing can
be done every 30 seconds and not be much of a bother to a fast processor.

Zero-Crossing Detectors Circuits and Applications

VFD phase loss fault

This (and a handful of other sources) suggests that phase loss prevention generally looks at ripple on the DC bus.

Even if the VFD was looking for voltage presence on each phase, I'm not sure that jumping it out would work. Most VFDs don't have a neutral so they would be testing each phase to the other two phases.
 
This suggests that it's typically done by measuring the ripple on the DC bus.

VFD phase loss fault


Yep, that is a good way, and it ends up looking for other faults as well, like failed capacitors, etc, etc. It does evade the usual "bridge two power terminals" solution, which will not help.

However, if you properly de-rate the drive, the ripple detector should NOT throw a fault. One big issue for derating is cap bank ripple current, and the ripple check will throw a fault if you go over the derated limit.
 
VFD phase loss fault

This (and a handful of other sources) suggests that phase loss prevention generally looks at ripple on the DC bus.

Even if the VFD was looking for voltage presence on each phase, I'm not sure that jumping it out would work. Most VFDs don't have a neutral so they would be testing each phase to the other two phases.

A capacitor on the jump would do it. A different way to measure from #16 is to use current transformers.
 








 
Back
Top