PDA

View Full Version : VFD single phase input?



ahidley
09-16-2015, 11:46 PM
I saw on another thread that using single phase input you recommended jumping L2 and L3. .si now I'm thinking "hmmmmm" I have a Mitsubishi vfd and I'm inputting 220single phase but this reduces capacity by 1/3 which is understandable. Can I jump one of the used inputs to the unused input? This would provide power to all three of the inputs but it would be out of phase. But would that matter because the vfd generates its own three phase output. So if this is possible then my vfd would nolonger be derated due to single phaseinput . Or is this only applicable to modern vfds? Mine is a Mitsubishi vfd z024. This is not mentioned in the manual.

J Tiers
09-16-2015, 11:59 PM
The benefit of the jumper is limited to fooling any phase loss detectors so the unit will run. no benefit in operation to speak of.

RB211
09-17-2015, 12:49 AM
It depends on your VFD, and who makes it, does it not? Mine is rated for single phase input, I use L1 and L3, I believe.

J Tiers
09-17-2015, 06:13 AM
If rated single phase in, then there is no reason to do any jumper unless so directed by the manual.

If not rated single phase, you may need it. And you should assume 50% derate unless otherwise stated. And there is a chance the VFD will not accept single phase even with ĵumper.

MaxHeadRoom
09-17-2015, 10:05 AM
I used to use Mitsubishi VFD's extensively, there is no benefit from jumpering as the input is simply a 3ph rectifier, there is however on some of the 1ph/3ph input types, two of the inputs are dedicated for the low voltage supply, so in 1ph use the power has to go on the recommended pair, on these models there is also an option of leaving the low voltage control on in standby and the main DC power off until ready for use.
Max.

lakeside53
09-17-2015, 11:20 AM
If rated single phase in, then there is no reason to do any jumper unless so directed by the manual.

If not rated single phase, you may need it. And you should assume 50% derate unless otherwise stated. And there is a chance the VFD will not accept single phase even with ĵumper.

What Jerry said! it's generally a myth that you can get 2/3 power out with one phase missing. 50% is a good assumption, but some can be way worse. In my case, I have pair of high end 7.5hp VFDs that are only rated at 1.5hp on single phase. My Hitachi are roughly 50%.

Rich Carlstedt
09-17-2015, 12:24 PM
I have wired many VFDs and built well over 50 rotary and static convertors, and have been interested in single phase to 3 phase changes for 50 years. A few years ago I had the chance to talk to a VFD Engineer and it was an enlightening experience. let me talk about 3 phase to 3 phase VFD's only here:

First, the input for each phase has a capacitor that stores the energy and provides power to a rectifier circuit which provides the power to the electronics that create 3 phase power. This electronic output could care less about 60 cycle current or incoming phases, it only wants a continuous source of DC power . If this level of incoming power is reduced, the VFD amperage output is reduced ( but voltage is maintained ).
When you provide only a single phase input, the capacitors in that 1/3 portion of the input determine the total output, based on THEIR capacity.
Some VFD builders have liberal caps used here, and some have only marginal capacity ...know that !
He also said that not all the caps are the same--they may be rated the same--but electronic components are notorious to vary +/-10 %, so you need to try all three input possibilities to find the best for output performance.
All 3 to 3 phase convertors will work on single phase input "IF" they do not have a voltage comparison circuit to prevent failure ( safety feature) . Most VFDS do have this feature however, but you may be able to disarm it.

I have to assume the jumping of the terminals is an attempt to increase capacity --maybe it works--maybe it doesn't --try it and see as VFD's vary in design

Rich

I hooked up a Hitachi 200 to a 7 Hp motor in a 15 inch Nardini Lathe.
The 10 HP VFD drives the 2 speed motor without issues and it has single phase input and
there was a noticeable difference in the input terminals used.
I have found Hitachi's to be the best in handling single phase loads

MaxHeadRoom
09-17-2015, 01:34 PM
I have to assume the jumping of the terminals is an attempt to increase capacity --maybe it works--maybe it doesn't --try it and see as VFD's vary in design



On all the VFD's I have stripped for parts, the majority have a 3ph full wave rectifier feeding a common capacitor bank for the DC. So strapping 2 phases effectively double up two of the rectifiers but has no effect on the DC.
When I asked a Mitsubishi engineer on the question of output of 3ph over 1ph he said that the rated Kw it can supply applies for both 1ph or 3ph.
Max.

J Tiers
09-17-2015, 02:21 PM
I have wired many VFDs and built well over 50 rotary and static convertors, and have been interested in single phase to 3 phase changes for 50 years. A few years ago I had the chance to talk to a VFD Engineer and it was an enlightening experience. let me talk about 3 phase to 3 phase VFD's only here:

First, the input for each phase has a capacitor that stores the energy and provides power to a rectifier circuit which provides the power to the electronics that create 3 phase power. This electronic output could care less about 60 cycle current or incoming phases, it only wants a continuous source of DC power . If this level of incoming power is reduced, the VFD amperage output is reduced ( but voltage is maintained ).
When you provide only a single phase input, the capacitors in that 1/3 portion of the input determine the total output, based on THEIR capacity.
Some VFD builders have liberal caps used here, and some have only marginal capacity ...know that !
He also said that not all the caps are the same--they may be rated the same--but electronic components are notorious to vary +/-10 %, so you need to try all three input possibilities to find the best for output performance.
All 3 to 3 phase convertors will work on single phase input "IF" they do not have a voltage comparison circuit to prevent failure ( safety feature) . Most VFDS do have this feature however, but you may be able to disarm it.

I have to assume the jumping of the terminals is an attempt to increase capacity --maybe it works--maybe it doesn't --try it and see as VFD's vary in design

Rich

I hooked up a Hitachi 200 to a 7 Hp motor in a 15 inch Nardini Lathe.
The 10 HP VFD drives the 2 speed motor without issues and it has single phase input and
there was a noticeable difference in the input terminals used.
I have found Hitachi's to be the best in handling single phase loads


nope.....

Same capacitor is used for ALL phases, that is not the issue.

The jumper may fool the unit into accepting all 3 phases as working. Done.

Limits are because 3 phase rectified with NO capacitor is good DC, only about 14% ripple. So the capacitor is just for a bypass, and may be very small. voltage never goes below about 86% of peak, and input current is quite steady.

With single phase, input drops to zero volts (!) twice per cycle. So the capacitor must be perhaps 100 x larger. AND input current is very "peaky". It is zero when onput voltage is below the DC, and high when the voltage is at peak, and the capacitor is being re-charged.

3 important limits are at work

Limit 1 is that the small 3phase capacitor may not hold voltage up high enough, so you get an undervoltage trip-off.

Limit 2 is that the single phase current peaks may be far more than the 3 phase rectifiers can take, because they were sized for 3 phase input current (lower).

Limit 3 is that the capacitor on 3 phase has to supply a lot of current when volts are low, and is charged with a high pulse of current when input volts are high. A capacitor that is selected for the small variation on 3 phase may not be able to take the higher current on single phase.

oxford
09-17-2015, 03:15 PM
I have a Toshiba(would have to look at the model) in my lathe. It does a little better than 50% de-rate for single in. I have a 5hp drive with a 3hp motor and no issues. I talked to a tech there before I got it and he told me a 3hp drive would be border line but would probably work for a 2hp motor with single line in.

PStechPaul
09-17-2015, 03:22 PM
Those are some very good points, and I have little to add. However:

1. Some VFDs have a transformer connected between two of the input phases, probably used for fans or control circuitry. It should be easy to tell by checking resistance on the input terminals. And a VFD like that cannot be run on DC unless you bypass the rectifiers and supply a separate AC source or DC for that purpose. I found that on a very old VFD that I have now stripped down, and modern units probably are not built that way.

http://enginuitysystems.com/pix/electrical/VFD_IDM_2013.jpg

The space in the corner is where this transformer and power supply was mounted:

http://enginuitysystems.com/pix/electrical/VFD_IDM_2032.jpg

2. What is needed to operate a VFD on single phase at full power is a PFC front end that is basically a boost switching regulator that uses most of the input waveform near the zero crossings so as not to need so much at the peaks to keep the main bus capacitors fully charged. Here is a schematic of a PFC circuit used in a 10 kW EV charger - the complete kit is less than $1500 so the PFC section probably could be built for under $500:

http://enginuitysystems.com/files/EMW/Driver_Board_Schematic_V14.png

Here is the power section:

http://enginuitysystems.com/files/EMW/Power.jpg

3. It might be possible to increase the power output by connecting a motor run capacitor from one of the single phase AC lines to the unused third phase, similar to how a 3 phase motor can be driven from single phase, and also similar to simple phase converters. The capacitor would need to be sized so that it would draw the full phase current at the input voltage and the power being delivered. So a 10 HP (7.5 kVA) VFD on 240 VAC would draw about (7500/240)/sqrt(3) or 18 amps which would be a reactance of 13 ohms and a capacitance of C=1/(2*pi*60*13) = 199 uF.

Sparky_NY
09-17-2015, 03:49 PM
3. It might be possible to increase the power output by connecting a motor run capacitor from one of the single phase AC lines to the unused third phase, similar to how a 3 phase motor can be driven from single phase, and also similar to simple phase converters. The capacitor would need to be sized so that it would draw the full phase current at the input voltage and the power being delivered. So a 10 HP (7.5 kVA) VFD on 240 VAC would draw about (7500/240)/sqrt(3) or 18 amps which would be a reactance of 13 ohms and a capacitance of C=1/(2*pi*60*13) = 199 uF.

I never tried a cap to feed the unused phase input but have always wondered if that wouldn't work fairly well. There should be about a 90 degree phase shift across the cap which would somewhat simulate 3 phase input but not at the balanced 120 degree phasing of true 3 phase. Its a interesting concept though and I have never seen it attempted.

Your calculation using sqrt(3) may not apply with the unusual phase relationship obtained with the cap method of generating a 3rd leg. Using sqrt(3) to calc a effective voltage is the standard for normal 3 phase with 120 degree phase relationships. I believe the phasing with a cap generated 3rd leg would be 0/90/180, do you agree? This would no doubt unbalance the current draws and invalidate the cap size calculation.

Another interesting consideration would be the ripple frequency on the VFD internal filter caps. The ripple would be 180hz with true 3 phase input, 120hz with single phase input. It makes my head hurt to contemplate what it would be with the cap generated 3rd leg and resulting strange phase relationship. It would have a effect on the VFD sensing ripple current though, and many VFD's use that for determining input phase loss.

I will be interested in your comments. I just bought a 7.5hp cnc lathe with a 5.5kw yaskawa vfd internal to it. I will never load the lathe up to its 7.5hp level but would sure like to run it off single phase. Alternately I will run the electronics off a 120 circuit and the VFD off a 15hp rotary phase converter. I wonder if the cap method would work for me, I will probably try it and see.

Rich Carlstedt
09-17-2015, 04:29 PM
nope.....

Same capacitor is used for ALL phases, that is not the issue.............

You missed it Jerry, not all VFD's have a single cap !
And that is the issue

The quality range of VFD's is immense
You can buy them with small caps, large caps or maybe with 'micro" caps you mentioned
The ability to convert power from one form or another is subject to the quality of the tool.

As I said , both from my experience and talking to others "you sometimes have to give it a try"
Telling anyone they can...or they can't.. is a fools game when you don't know what they have for a tool

My best
Rich

Sparky_NY
09-17-2015, 04:35 PM
You missed it Jerry, not all VFD's have a single cap !
And that is the issue

The quality range of VFD's is immense
You can buy them with small caps, large caps or maybe with 'micro" caps you mentioned
The ability to convert power from one form or another is subject to the quality of the tool.

As I said , both from my experience and talking to others "you sometimes have to give it a try"
Telling anyone they can...or they can't.. is a fools game when you don't know what they have for a tool

My best
Rich

I don't think Jerry missed it. Yes there can be many types and sizes of caps but they form a single filter bank. Your original post claims the caps feed the rectifiers, that is way wrong, exactly reversed. They are configured as a standard 3 phase line fed rectifer followed by a standard bank of filter caps to form the DC bus. Single phase rated VFD's would have much higher value filtering caps due to the lower ripple frequency.

PStechPaul
09-17-2015, 04:50 PM
In the case of a motor, or a balanced resistive/inductive load, the phase of the voltage on the third leg will depend on the complex impedance of that point. The capacitor will supply current at 90 degree phase to that of L1-L2, but the load, which appears as resistance, capacitance, and inductance across all three phases, cause an additional phase shift. If you established a center point between L1 and L2, where single phase is applied, you could generate a third phase at 90 degrees that would form the third phase on L3. The amplitude would be 240*sin(60) or 208 VAC. The neutral of the resulting three phase power would be at 240*sqrt(3) or 138V to each phase. So I am not sure if my value for the capacitance is correct. However, it should be "in the ballpark". I will try a simulation to see how things look. It is difficult to simulate a VFD and a motor, but an ideal motor should be close to a set of three resistors with a small amount of series inductance for a motor with 95% or so PF running a constant full load. I think a three phase motor actually draws a constant current from the DC bus. So it will be interesting to see the input waveforms depending on capacitor size and load.

J Tiers
09-17-2015, 04:50 PM
Never seen one with multiple banks, and I am not sure what the advantage would be. It never occurred to me to design one like that, no need. Not to say there may not BE one, or two that are like that, a delta configured unit would necessarily be made that way, but there is no good reason to make one that way for general use.

And it would be expensive, not a good thing in a competitive market.

The vast majority are made as described. One bank of capacitors, configured in parallel as one large unit

As for capacitor phase shifting, that may not work at all. The shift is dependent on the load impedance, and with a rectifier, the effective impedance varies more than 100:1 as the rectifiers conduct or block.

Sparky_NY
09-17-2015, 04:59 PM
In the case of a motor, or a balanced resistive/inductive load, the phase of the voltage on the third leg will depend on the complex impedance of that point. The capacitor will supply current at 90 degree phase to that of L1-L2, but the load, which appears as resistance, capacitance, and inductance across all three phases, cause an additional phase shift. If you established a center point between L1 and L2, where single phase is applied, you could generate a third phase at 90 degrees that would form the third phase on L3. The amplitude would be 240*sin(60) or 208 VAC. The neutral of the resulting three phase power would be at 240*sqrt(3) or 138V to each phase. So I am not sure if my value for the capacitance is correct. However, it should be "in the ballpark". I will try a simulation to see how things look. It is difficult to simulate a VFD and a motor, but an ideal motor should be close to a set of three resistors with a small amount of series inductance for a motor with 95% or so PF running a constant full load. I think a three phase motor actually draws a constant current from the DC bus. So it will be interesting to see the input waveforms depending on capacitor size and load.

I had to read that about 3X to digest it ! LOL I am thinking the theory is sound, the actual capacitor value may have to be determined by experiment, much like a rotary phase converter run caps. To further thicken the plot, the VFD motor is not always fully loaded so input current to the VFD will be a range rather than a fixed value. That means the ideal cap value will vary with motor load. Too many unknown variables ! Still, the theory is sound and feeding the 3rd input leg with a cap generated supply would help share the line load between all 3 legs of the input rectifier bank, even if not well balanced its still far better than using only 2 rectifier input legs. I believe filter cap ripple would also benefit.

Interesting discussion !

Sparky_NY
09-17-2015, 05:06 PM
As for capacitor phase shifting, that may not work at all. The shift is dependent on the load impedance, and with a rectifier, the effective impedance varies more than 100:1 as the rectifiers conduct or block.

Ah yes, yet one more factor that comes into play ! I cannot see any possible harm in experimenting with it a bit. Any load shifted to the previously unused rectifier input will lighten the load on the remaining two that are directly line connected.

Also, I am wondering what difference it makes what the phase shift across the cap is when the rectifier is not conducting anyways?

Rich Carlstedt
09-17-2015, 05:32 PM
I don't think Jerry missed it. Yes there can be many types and sizes of caps but they form a single filter bank. Your original post claims the caps feed the rectifiers, that is way wrong, exactly reversed. They are configured as a standard 3 phase line fed rectifer followed by a standard bank of filter caps to form the DC bus. Single phase rated VFD's would have much higher value filtering caps due to the lower ripple frequency.

If it came out that way, it's not what I meant. Each leg feeds a rectifier and that has a cap ( storing the output). Three legs = 3 caps that fed the invertor section
In single phase input, only one cap is needed
When you have one Cap fed by a three legged bridge as Jerry pointed out, then that cap is maxed out. I understand.
The EE that I spoke to, and some of the VFD's i have worked on had 3 caps ( like a Parajust unit ) .

I have no argument about making them cheaper....as Jerry mentioned.
If you have a 3 phase input unit, and need single phase, then just kitbash it and use all three rectifiers as a common bridge and add 37 % or more capacity in the cap to handle ripple.
Just a thought !
The only reason you downgrade the performance is a lack of rectifier current and capacitor ratings
Rich

J Tiers
09-17-2015, 06:47 PM
Really, the use of multiple cap banks for different phases is rare enough that it is pretty fair to say "nobody here will see one". Mentioning them at all is basically "for completeness", not as a practical matter when dealing with units to 5 HP at 230V.

You almost certainly will not see one from TECO, Hitachi, Invertek, TB Woods, Vacon, etc etc at our power levels.

As for the ones actually made like that, unless you are dealing with specialized industrial uses, "you will never see one" is a fair statement.

Macona finds odd stuff, maybe he might see one.

kf2qd
09-17-2015, 08:25 PM
I have seen VFDs and Servo drives that are listed for 3Ph/1ph and DO jumper L2 & L3 when used with single phase. (Yaskawa drives)

SO - instead of asking us here how to wire it, you would probably get much better advice by going to the manufacturers data. I know it can be hard for us to assume the manufacturer actually knew what he was doing, and that somebody we only know through this forum might not be the best source for information...

Sparky_NY
09-17-2015, 08:50 PM
I have seen VFDs and Servo drives that are listed for 3Ph/1ph and DO jumper L2 & L3 when used with single phase. (Yaskawa drives)

SO - instead of asking us here how to wire it, you would probably get much better advice by going to the manufacturers data. I know it can be hard for us to assume the manufacturer actually knew what he was doing, and that somebody we only know through this forum might not be the best source for information...

Many (most) of the 3 phase drives work nicely on single phase provided they are derated. The manufacturers usually do not condone this type of operation and as a result they do not provide support nor documentation for this type of operation in many cases. So, chances are you are not going to find anything in their data that will be of any use. Still, it is very common practice and works well providing the derating is observed.

lakeside53
09-17-2015, 09:30 PM
Here's Hitachi's generalized info :

http://www.hitachi-america.us/supportingdocs/forbus/inverters/Support/AN032404-1_Rev_A_Sizing_for_Single-Phase.pdf

Basically.. select a VFD double the name plate current. They say unless it's to correct a phase detect issue, the unused input should remain unconnected.

macona
09-17-2015, 09:44 PM
A lot of VFDs, especially the larger ones have terminals for the DC bus. You could add more bulk capacitance there. I did this once on an allen bradley servo drive to help with a regen over voltage issue.

J Tiers
09-17-2015, 09:57 PM
A lot of VFDs, especially the larger ones have terminals for the DC bus. You could add more bulk capacitance there. I did this once on an allen bradley servo drive to help with a regen over voltage issue.

You can do that.

You might have trouble with burning turn-on surge resistors, if the capacitor is too big, and takes too long to charge.

PStechPaul
09-17-2015, 10:54 PM
I did a simulation of the capacitor to the third leg, and it really doesn't do much of anything. It does seem to work to some extent into three resistive/inductive loads tied in delta, but not at the input to a three phase rectifier bridge. So probably the old "trick" of adding a three phase idler motor, perhaps with such a capacitor, may work.

But I think the ideal solution (other than a Phase Perfect) is the PFC front end. If you can generate 380 VDC from 240 VAC at 95% PF then you can just bypass the rectifiers and connect directly to the DC bus link. You do need to be careful when first applying power, as the boost converter does not limit current to the output capacitors. I have made a surge limiter that seems to work well in simulation:

http://enginuitysystems.com/files/EMW/120SIne-320DC_Doubler_Surge_Discharge.png

Sparky_NY
09-17-2015, 10:55 PM
A lot of VFDs, especially the larger ones have terminals for the DC bus. You could add more bulk capacitance there. I did this once on an allen bradley servo drive to help with a regen over voltage issue.

Great idea ! I have a cnc lathe that I just bought with a 7.6hp spindle and a 5.5kw yaskawa vfd. I sure would like to get away with running it on single phase. I will never use anywhere near the full hp of the lathe but still question if I could get away with that VFD on single phase. Maybe with some external C added like you mentioned ! (and lighten up the accel setting)

macona
09-18-2015, 03:40 AM
Chances are you will be able to run that spindle just fine on single phase with no mods. Unless you are doing some real hard acceleration and heavy turning you will never see an issue.

In my Monarch I use a Mitsubishi 5kw servo motor for the spindle and I have never had an issue with power. Remember, Japanese made devices are made to run on 200-240v 50/60hz with usually a +/-10% tolerance so that means they should be able to run fine on 180v 50hz, put 240 60hz and you probably will never see an under voltage bus fault. I hit motor torque limits before I have ever hit that error.

Sparky_NY
09-18-2015, 06:22 AM
Chances are you will be able to run that spindle just fine on single phase with no mods. Unless you are doing some real hard acceleration and heavy turning you will never see an issue.

In my Monarch I use a Mitsubishi 5kw servo motor for the spindle and I have never had an issue with power. Remember, Japanese made devices are made to run on 200-240v 50/60hz with usually a +/-10% tolerance so that means they should be able to run fine on 180v 50hz, put 240 60hz and you probably will never see an under voltage bus fault. I hit motor torque limits before I have ever hit that error.

Music to my ears ! A few years back I had a 3hp ABB VFD on a bridgeport (2HP), 3ph rated only and it would trip off with a "high ripple voltage" fault. This motor/VFD is matched in size so I had had doubts. I am now a lot more optimistic.

J Tiers
09-18-2015, 08:51 AM
You can add your own rectifier and cap if external dc terminals are available, AND AC is not used for anything in the unit. You then have a "system drive" equivalent.

If it senses AC or has an ac power supply for controls, etc, you will have to deal with that. And you may need your own inrush protector in any case.

Sparky_NY
09-18-2015, 09:16 AM
In light of Macona's statements and keeping with the old principle of not fixing it if it isn't broken, I will have to give it a try and see what happens. I have a plan B, a 15hp rotophase I can use for 3ph if needed and just run the control/drives etc off a independent single phase circuit. Sure would be nice not to have to do any of that stuff though, time will tell.

ahidley
09-20-2015, 01:52 AM
Ok. I put a jumper on. Mine only faults out once in a while. If it does I'll post back here. Thanks